feat(repo-map): extract clean repo map branch

2026-04-14 18:57:46 +05:30
114 changed files with 2786 additions and 7385 deletions
--- a/.env.example
+++ b/.env.example
@@ -225,30 +225,6 @@ ANTHROPIC_API_KEY=sk-ant-your-key-here
 # GOOGLE_CLOUD_PROJECT=your-gcp-project-id


-# -----------------------------------------------------------------------------
-# Option 9: NVIDIA NIM
-# -----------------------------------------------------------------------------
-# NVIDIA NIM provides hosted inference endpoints for NVIDIA models.
-# Get your API key from https://build.nvidia.com/
-#
-# CLAUDE_CODE_USE_OPENAI=1
-# NVIDIA_API_KEY=nvapi-your-key-here
-# OPENAI_BASE_URL=https://integrate.api.nvidia.com/v1
-# OPENAI_MODEL=nvidia/llama-3.1-nemotron-70b-instruct
-
-
-# -----------------------------------------------------------------------------
-# Option 10: MiniMax
-# -----------------------------------------------------------------------------
-# MiniMax API provides text generation models.
-# Get your API key from https://platform.minimax.io/
-#
-# CLAUDE_CODE_USE_OPENAI=1
-# MINIMAX_API_KEY=your-minimax-key-here
-# OPENAI_BASE_URL=https://api.minimax.io/v1
-# OPENAI_MODEL=MiniMax-M2.5
-
-
 # =============================================================================
 # OPTIONAL TUNING
 # =============================================================================
--- a/.github/workflows/release.yml
+++ b/.github/workflows/release.yml
@@ -11,7 +11,6 @@ concurrency:

 jobs:
  release-please:
-    if: ${{ github.repository == 'Gitlawb/openclaude' }}
    name: Release Please
    runs-on: ubuntu-latest
    permissions:
--- a/.release-please-manifest.json
+++ b/.release-please-manifest.json
@@ -1,3 +1,3 @@
 {
-  ".": "0.5.0"
+  ".": "0.3.0"
 }
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,43 +1,5 @@
 # Changelog

-## [0.5.0](https://github.com/Gitlawb/openclaude/compare/v0.4.0...v0.5.0) (2026-04-20)
-
-
-### Features
-
-* add OPENCLAUDE_DISABLE_STRICT_TOOLS env var to opt out of strict MCP tool schema normalization ([#770](https://github.com/Gitlawb/openclaude/issues/770)) ([e6e8d9a](https://github.com/Gitlawb/openclaude/commit/e6e8d9a24897e4c9ef08b72df20fabbf8ef27f38))
-* mask provider api key input ([#772](https://github.com/Gitlawb/openclaude/issues/772)) ([13e9f22](https://github.com/Gitlawb/openclaude/commit/13e9f22a83a2b0f85f557b1e12c9442ba61241e4))
-
-
-### Bug Fixes
-
-* allow provider recovery during startup ([#765](https://github.com/Gitlawb/openclaude/issues/765)) ([f828171](https://github.com/Gitlawb/openclaude/commit/f828171ef1ab94e2acf73a28a292799e4e26cc0d))
-* **api:** drop orphan tool results to satisfy strict role sequence ([#745](https://github.com/Gitlawb/openclaude/issues/745)) ([b786b76](https://github.com/Gitlawb/openclaude/commit/b786b765f01f392652eaf28ed3579a96b7260a53))
-* **help:** prevent /help tab crash from undefined descriptions ([#732](https://github.com/Gitlawb/openclaude/issues/732)) ([3d1979f](https://github.com/Gitlawb/openclaude/commit/3d1979ff066db32415e0c8321af916d81f5f2621))
-* **mcp:** sync required array with properties in tool schemas ([#754](https://github.com/Gitlawb/openclaude/issues/754)) ([002a8f1](https://github.com/Gitlawb/openclaude/commit/002a8f1f6de2fcfc917165d828501d3047bad61f))
-* remove cached mcpClient in diagnostic tracking to prevent stale references ([#727](https://github.com/Gitlawb/openclaude/issues/727)) ([2c98be7](https://github.com/Gitlawb/openclaude/commit/2c98be700274a4241963b5f43530bf3bd8f8963f))
-* use raw context window for auto-compact percentage display ([#748](https://github.com/Gitlawb/openclaude/issues/748)) ([55c5f26](https://github.com/Gitlawb/openclaude/commit/55c5f262a9a5a8be0aa9ae8dc6c7dafc465eb2c6))
-
-## [0.4.0](https://github.com/Gitlawb/openclaude/compare/v0.3.0...v0.4.0) (2026-04-17)
-
-
-### Features
-
-* add Alibaba Coding Plan (DashScope) provider support ([#509](https://github.com/Gitlawb/openclaude/issues/509)) ([43ac6db](https://github.com/Gitlawb/openclaude/commit/43ac6dba75537282da1e2ad8f855082bc4e25f1e))
-* add NVIDIA NIM and MiniMax provider support ([#552](https://github.com/Gitlawb/openclaude/issues/552)) ([51191d6](https://github.com/Gitlawb/openclaude/commit/51191d61326e1f8319d70b3a3c0d9229e185a564))
-* add ripgrep to Dockerfile for faster file searching ([#688](https://github.com/Gitlawb/openclaude/issues/688)) ([12dd375](https://github.com/Gitlawb/openclaude/commit/12dd3755c619cc27af3b151ae8fdb9d425a7b9a2))
-* **api:** classify openai-compatible provider failures ([#708](https://github.com/Gitlawb/openclaude/issues/708)) ([80a00ac](https://github.com/Gitlawb/openclaude/commit/80a00acc2c6dc4657a78de7366f7a9ebc920bfbb))
-* **vscode:** add full chat interface to OpenClaude extension ([#608](https://github.com/Gitlawb/openclaude/issues/608)) ([fbcd928](https://github.com/Gitlawb/openclaude/commit/fbcd928f7f8511da795aea3ad318bddf0ab9a1a7))
-
-
-### Bug Fixes
-
-* focus "Done" option after completing provider manager actions ([#718](https://github.com/Gitlawb/openclaude/issues/718)) ([d6f5130](https://github.com/Gitlawb/openclaude/commit/d6f5130c204d8ffe582212466768706cd7fd6774))
-* **models:** prevent /models crash from non-string saved model values ([#691](https://github.com/Gitlawb/openclaude/issues/691)) ([6b2121d](https://github.com/Gitlawb/openclaude/commit/6b2121da12189fa7ce1f33394d18abd24cf8a01b))
-* prevent crash in commands tab when description is undefined ([#730](https://github.com/Gitlawb/openclaude/issues/730)) ([eed77e6](https://github.com/Gitlawb/openclaude/commit/eed77e6579866a98384dcc948a0ad6406614ede3))
-* strip comments before scanning for missing imports ([#676](https://github.com/Gitlawb/openclaude/issues/676)) ([a00b792](https://github.com/Gitlawb/openclaude/commit/a00b7928de9662ffb7ef6abd8cd040afe6f4f122))
-* **ui:** show correct endpoint URL in intro screen for custom Anthropic endpoints ([#735](https://github.com/Gitlawb/openclaude/issues/735)) ([3424663](https://github.com/Gitlawb/openclaude/commit/34246635fb9a09499047a52e7f96ca9b36c8a85a))
-
 ## [0.3.0](https://github.com/Gitlawb/openclaude/compare/v0.2.3...v0.3.0) (2026-04-14)


--- a/9
+++ b/9
@@ -36,11 +36,14 @@ COPY --from=build /app/node_modules/ node_modules/
 COPY --from=build /app/package.json package.json
 COPY README.md ./

-# Install git and ripgrep — many CLI tool operations depend on them
-RUN apt-get update && apt-get install -y --no-install-recommends git ripgrep \
+# Install git — many CLI tool operations depend on it
+RUN apt-get update && apt-get install -y --no-install-recommends git \
    && rm -rf /var/lib/apt/lists/*

 # Run as non-root user
-USER node
+RUN groupadd --gid 1000 appuser && useradd --uid 1000 --gid appuser --shell /bin/bash --create-home appuser
+USER appuser
+WORKDIR /home/appuser
+ENV HOME=/home/appuser

 ENTRYPOINT ["node", "/app/dist/cli.mjs"]
--- a/README.md
+++ b/README.md
@@ -15,10 +15,6 @@ OpenClaude is also mirrored to GitLawb:

 [Quick Start](#quick-start) | [Setup Guides](#setup-guides) | [Providers](#supported-providers) | [Source Build](#source-build-and-local-development) | [VS Code Extension](#vs-code-extension) | [Community](#community)

-## Star History
-
-[![Star History Chart](https://api.star-history.com/chart?repos=gitlawb/openclaude&type=date&legend=top-left)](https://www.star-history.com/?repos=gitlawb%2Fopenclaude&type=date&legend=top-left)
-
 ## Why OpenClaude

 - Use one CLI across cloud APIs and local model backends
@@ -92,16 +88,6 @@ $env:OPENAI_MODEL="qwen2.5-coder:7b"
 openclaude
 ```

-### Using Ollama's launch command
-
-If you have [Ollama](https://ollama.com) installed, you can skip the env var setup entirely:
-
-```bash
-ollama launch openclaude --model qwen2.5-coder:7b
-```
-
-This automatically sets `ANTHROPIC_BASE_URL`, model routing, and auth so all API traffic goes through your local Ollama instance. Works with any model you have pulled — local or cloud.
-
 ## Setup Guides

 Beginner-friendly guides:
@@ -124,7 +110,7 @@ Advanced and source-build guides:
 | GitHub Models | `/onboard-github` | Interactive onboarding with saved credentials |
 | Codex OAuth | `/provider` | Opens ChatGPT sign-in in your browser and stores Codex credentials securely |
 | Codex | `/provider` | Uses existing Codex CLI auth, OpenClaude secure storage, or env credentials |
-| Ollama | `/provider`, env vars, or `ollama launch` | Local inference with no API key |
+| Ollama | `/provider` or env vars | Local inference with no API key |
 | Atomic Chat | advanced setup | Local Apple Silicon backend |
 | Bedrock / Vertex / Foundry | env vars | Additional provider integrations for supported environments |

@@ -331,8 +317,7 @@ For larger changes, open an issue first so the scope is clear before implementat
 - `bun run build`
 - `bun run test:coverage`
 - `bun run smoke`
- focused `bun test ...` runs for files and flows you changed
-
+- focused `bun test ...` runs for touched areas

 ## Disclaimer

--- a/docs/advanced-setup.md
+++ b/docs/advanced-setup.md
@@ -84,16 +84,6 @@ OpenRouter model availability changes over time. If a model stops working, try a

 ### Ollama

-Using `ollama launch` (recommended if you have Ollama installed):
-
-```bash
-ollama launch openclaude --model llama3.3:70b
-```
-
-This handles all environment setup automatically — no env vars needed. Works with any local or cloud model available in your Ollama instance.
-
-Using environment variables manually:
-
 ```bash
 ollama pull llama3.3:70b

--- a/docs/repo-map.md
+++ b/docs/repo-map.md
@@ -0,0 +1,67 @@
+# Codebase Intelligence — Repo Map
+
+The repo map feature gives the AI model structural awareness of your codebase at the start of each session. Instead of the model needing to explore the repository with `Grep`, `Glob`, and `Read` calls, it starts with a ranked summary of the most important files and their key signatures.
+
+## How it works
+
+1. **File enumeration** — Lists all tracked files via `git ls-files` (falls back to a manual directory walk when not in a git repo)
+2. **Symbol extraction** — Parses each supported source file with tree-sitter to extract function, class, type, and interface definitions, plus cross-file references
+3. **Reference graph** — Builds a directed graph where an edge from file A to file B means A references a symbol defined in B. Edges are weighted by reference count multiplied by the IDF (inverse document frequency) of the symbol name — common names like `get`, `set`, `value` contribute less
+4. **PageRank** — Ranks files by structural importance using PageRank. Files imported by many others rank highest
+5. **Rendering** — Walks ranked files top-down, emitting file paths and definition signatures, stopping when the token budget is reached
+
+Results are cached to disk (`~/.openclaude/repomap-cache/`) keyed by file path, mtime, and size. Only changed files are re-parsed on subsequent runs.
+
+## Supported languages
+
+- TypeScript (`.ts`, `.tsx`)
+- JavaScript (`.js`, `.jsx`, `.mjs`, `.cjs`)
+- Python (`.py`)
+
+Additional language grammars will be added in future releases.
+
+## Enabling auto-injection
+
+The repo map is gated behind the `REPO_MAP` feature flag, **off by default**. To enable auto-injection into the session context:
+
+Set the environment variable before launching:
+
+```bash
+REPO_MAP=1 openclaude
+```
+
+Or add it to your shell profile for persistent use.
+
+When enabled, the map is built once per session and prepended to the system context alongside git status and CLAUDE.md content. The default budget is 1024 tokens.
+
+Auto-injection is skipped in:
+- Bare mode (`--bare`)
+- Remote sessions (`CLAUDE_CODE_REMOTE`)
+
+## The /repomap slash command
+
+The `/repomap` command is always available regardless of the feature flag. It lets you inspect and tune the map interactively.
+
+```
+/repomap                          # Show the map with default settings (1024 tokens)
+/repomap --tokens 4096            # Increase the token budget for a larger map
+/repomap --focus src/tools/       # Boost specific paths in the ranking
+/repomap --focus src/context.ts   # Can use multiple --focus flags
+/repomap --stats                  # Show cache statistics
+/repomap --invalidate             # Clear cache and rebuild from scratch
+```
+
+## The RepoMap tool
+
+The model can also call the `RepoMap` tool on demand during a session. This is useful when:
+- The model needs structural context mid-conversation
+- The user asks about specific areas (the model can pass `focus_files` or `focus_symbols`)
+- A larger token budget is needed than the auto-injected default
+
+## Known limitations
+
+- **Signatures only** — The map shows function/class/type declarations, not implementations. The model still needs `Read` to see function bodies.
+- **Cold build time** — First build on large repos (2000+ files) can take 20-30 seconds due to WASM-based parsing. Subsequent builds use the disk cache and complete in under 100ms.
+- **Language coverage** — Only TypeScript, JavaScript, and Python are supported. Files in other languages are skipped.
+- **TypeScript references** — The TypeScript tree-sitter query captures type annotations and `new` expressions as references, but not plain function calls. This means the ranking slightly favors type-heavy hub files.
+- **Git dependency** — File enumeration uses `git ls-files` by default. Non-git repos fall back to a directory walk with hardcoded exclusions.
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
  "name": "@gitlawb/openclaude",
-  "version": "0.5.0",
+  "version": "0.3.0",
  "description": "Claude Code opened to any LLM — OpenAI, Gemini, DeepSeek, Ollama, and 200+ models",
  "type": "module",
  "bin": {
@@ -95,8 +95,12 @@
    "fuse.js": "7.1.0",
    "get-east-asian-width": "1.5.0",
    "google-auth-library": "9.15.1",
+    "graphology": "^0.26.0",
+    "graphology-operators": "^1.6.0",
+    "graphology-pagerank": "^1.1.0",
    "https-proxy-agent": "7.0.6",
    "ignore": "7.0.5",
+    "js-tiktoken": "^1.0.16",
    "indent-string": "5.0.0",
    "jsonc-parser": "3.3.1",
    "lodash-es": "4.18.1",
@@ -117,11 +121,13 @@
    "strip-ansi": "7.2.0",
    "supports-hyperlinks": "3.2.0",
    "tree-kill": "1.2.2",
+    "tree-sitter-wasms": "^0.1.12",
    "turndown": "7.2.2",
    "type-fest": "4.41.0",
    "undici": "7.24.6",
    "usehooks-ts": "3.1.1",
    "vscode-languageserver-protocol": "3.17.5",
+    "web-tree-sitter": "^0.25.0",
    "wrap-ansi": "9.0.2",
    "ws": "8.20.0",
    "xss": "1.0.15",
--- a/scripts/build.ts
+++ b/scripts/build.ts
@@ -367,17 +367,9 @@ export const SeverityNumber = {};
              const full = pathMod.join(dir, ent.name)
              if (ent.isDirectory()) { walk(full); continue }
              if (!/\.(ts|tsx)$/.test(ent.name)) continue
-              const rawCode: string = fs.readFileSync(full, 'utf-8')
+              const code: string = fs.readFileSync(full, 'utf-8')
              const fileDir = pathMod.dirname(full)

-              // Strip comments before scanning for imports/requires.
-              // The regex scanner matches require()/import() patterns
-              // inside JSDoc comments, causing false-positive missing
-              // module detection that breaks the build with noop stubs.
-              const code = rawCode
-                .replace(/\/\*[\s\S]*?\*\//g, '')  // block comments
-                .replace(/\/\/.*$/gm, '')           // line comments
-
              // Collect static imports: import { X } from '...'
              for (const m of code.matchAll(/import\s+(?:\{([^}]*)\}|(\w+))?\s*(?:,\s*\{([^}]*)\})?\s*from\s+['"](.*?)['"]/g)) {
                checkAndRegister(m[4], fileDir, m[1] || m[3] || '')
--- a/src/cli/handlers/mcp.tsx
+++ b/src/cli/handlers/mcp.tsx
@@ -11,12 +11,7 @@ import { MCPServerDesktopImportDialog } from '../../components/MCPServerDesktopI
 import { render } from '../../ink.js';
 import { KeybindingSetup } from '../../keybindings/KeybindingProviderSetup.js';
 import { type AnalyticsMetadata_I_VERIFIED_THIS_IS_NOT_CODE_OR_FILEPATHS, logEvent } from '../../services/analytics/index.js';
-import {
-  clearMcpClientConfig,
-  clearServerTokensFromSecureStorage,
-  readClientSecret,
-  saveMcpClientSecret,
-} from '../../services/mcp/auth.js'
+import { clearMcpClientConfig, clearServerTokensFromLocalStorage, readClientSecret, saveMcpClientSecret } from '../../services/mcp/auth.js';
 import { doctorAllServers, doctorServer, type McpDoctorReport, type McpDoctorScopeFilter } from '../../services/mcp/doctor.js';
 import { connectToServer, getMcpServerConnectionBatchSize } from '../../services/mcp/client.js';
 import { addMcpConfig, getAllMcpConfigs, getMcpConfigByName, getMcpConfigsByScope, removeMcpConfig } from '../../services/mcp/config.js';
--- a/src/commands.test.ts
+++ b/src/commands.test.ts
@@ -1,30 +0,0 @@
-import { formatDescriptionWithSource } from './commands.js'
-
-describe('formatDescriptionWithSource', () => {
-  test('returns empty text for prompt commands missing a description', () => {
-    const command = {
-      name: 'example',
-      type: 'prompt',
-      source: 'builtin',
-      description: undefined,
-    } as any
-
-    expect(formatDescriptionWithSource(command)).toBe('')
-  })
-
-  test('formats plugin commands with missing description safely', () => {
-    const command = {
-      name: 'example',
-      type: 'prompt',
-      source: 'plugin',
-      description: undefined,
-      pluginInfo: {
-        pluginManifest: {
-          name: 'MyPlugin',
-        },
-      },
-    } as any
-
-    expect(formatDescriptionWithSource(command)).toBe('(MyPlugin) ')
-  })
-})
--- a/src/commands.ts
+++ b/src/commands.ts
@@ -22,6 +22,7 @@ import ctx_viz from './commands/ctx_viz/index.js'
 import doctor from './commands/doctor/index.js'
 import onboardGithub from './commands/onboard-github/index.js'
 import memory from './commands/memory/index.js'
+import repomap from './commands/repomap/index.js'
 import help from './commands/help/index.js'
 import ide from './commands/ide/index.js'
 import init from './commands/init.js'
@@ -307,6 +308,7 @@ const COMMANDS = memoize((): Command[] => [
  releaseNotes,
  reloadPlugins,
  rename,
+  repomap,
  resume,
  session,
  skills,
@@ -740,23 +742,23 @@ export function getCommand(commandName: string, commands: Command[]): Command {
 */
 export function formatDescriptionWithSource(cmd: Command): string {
  if (cmd.type !== 'prompt') {
-    return cmd.description ?? ''
+    return cmd.description
  }

  if (cmd.kind === 'workflow') {
-    return `${cmd.description ?? ''} (workflow)`
+    return `${cmd.description} (workflow)`
  }

  if (cmd.source === 'plugin') {
    const pluginName = cmd.pluginInfo?.pluginManifest.name
    if (pluginName) {
-      return `(${pluginName}) ${cmd.description ?? ''}`
+      return `(${pluginName}) ${cmd.description}`
    }
-    return `${cmd.description ?? ''} (plugin)`
+    return `${cmd.description} (plugin)`
  }

  if (cmd.source === 'builtin' || cmd.source === 'mcp') {
-    return cmd.description ?? ''
+    return cmd.description
  }

  if (cmd.source === 'bundled') {
--- a/src/commands/provider/provider.test.tsx
+++ b/src/commands/provider/provider.test.tsx
@@ -401,7 +401,7 @@ test('buildCodexProfileEnv derives oauth source from secure storage when no expl
  })
 })

-test('explicitly declared env takes precedence over applySavedProfileToCurrentSession', async () => {
+test('applySavedProfileToCurrentSession switches the current env to the saved Codex profile', async () => {
  // @ts-expect-error cache-busting query string for Bun module mocks
  const { applySavedProfileToCurrentSession } = await import(
    '../../utils/providerProfile.js?apply-saved-profile-codex'
@@ -430,18 +430,18 @@ test('explicitly declared env takes precedence over applySavedProfileToCurrentSe

  expect(warning).toBeNull()
  expect(processEnv.CLAUDE_CODE_USE_OPENAI).toBe('1')
-  expect(processEnv.OPENAI_MODEL).toBe('gpt-4o')
+  expect(processEnv.OPENAI_MODEL).toBe('codexplan')
  expect(processEnv.OPENAI_BASE_URL).toBe(
-    "https://api.openai.com/v1",
+    'https://chatgpt.com/backend-api/codex',
  )
-  expect(processEnv.CODEX_API_KEY).toBeUndefined()
-  expect(processEnv.CHATGPT_ACCOUNT_ID).toBeUndefined()
-  expect(processEnv.OPENAI_API_KEY).toBe("sk-openai")
+  expect(processEnv.CODEX_API_KEY).toBe('codex-live')
+  expect(processEnv.CHATGPT_ACCOUNT_ID).toBe('acct_codex')
+  expect(processEnv.OPENAI_API_KEY).toBeUndefined()
  expect(processEnv.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED).toBeUndefined()
  expect(processEnv.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED_ID).toBeUndefined()
 })

-test('explicitly declared env takes precedence over applySavedProfileToCurrentSession', async () => {
+test('applySavedProfileToCurrentSession ignores stale Codex env overrides for OAuth-backed profiles', async () => {
  // @ts-expect-error cache-busting query string for Bun module mocks
  const { applySavedProfileToCurrentSession } = await import(
    '../../utils/providerProfile.js?apply-saved-profile-codex-oauth'
@@ -465,13 +465,13 @@ test('explicitly declared env takes precedence over applySavedProfileToCurrentSe
    processEnv,
  })

-  expect(warning).not.toBeUndefined()
-  expect(processEnv.OPENAI_MODEL).toBe('gpt-4o')
+  expect(warning).toBeNull()
+  expect(processEnv.OPENAI_MODEL).toBe('codexplan')
  expect(processEnv.OPENAI_BASE_URL).toBe(
-    "https://api.openai.com/v1",
+    'https://chatgpt.com/backend-api/codex',
  )
-  expect(processEnv.CODEX_API_KEY).toBe("stale-codex-key")
-  expect(processEnv.CHATGPT_ACCOUNT_ID).toBe('acct_stale')
+  expect(processEnv.CODEX_API_KEY).toBeUndefined()
+  expect(processEnv.CHATGPT_ACCOUNT_ID).not.toBe('acct_stale')
  expect(processEnv.CHATGPT_ACCOUNT_ID).toBeTruthy()
 })

@@ -487,8 +487,8 @@ test('buildCurrentProviderSummary redacts poisoned model and endpoint values', (
  })

  expect(summary.providerLabel).toBe('OpenAI-compatible')
-  expect(summary.modelLabel).toBe('sk-...678')
-  expect(summary.endpointLabel).toBe('sk-...678')
+  expect(summary.modelLabel).toBe('sk-...5678')
+  expect(summary.endpointLabel).toBe('sk-...5678')
 })

 test('buildCurrentProviderSummary labels generic local openai-compatible providers', () => {
--- a/src/commands/repomap/index.ts
+++ b/src/commands/repomap/index.ts
@@ -0,0 +1,17 @@
+/**
+ * /repomap command - minimal metadata only.
+ * Implementation is lazy-loaded from repomap.ts to reduce startup time.
+ */
+import type { Command } from '../../commands.js'
+
+const repomap = {
+  type: 'local',
+  name: 'repomap',
+  description:
+    'Show or configure the repository structural map (codebase intelligence)',
+  isHidden: false,
+  supportsNonInteractive: true,
+  load: () => import('./repomap.js'),
+} satisfies Command
+
+export default repomap
--- a/src/commands/repomap/repomap.test.ts
+++ b/src/commands/repomap/repomap.test.ts
@@ -0,0 +1,56 @@
+import { describe, expect, test } from 'bun:test'
+import { parseArgs } from './repomap.js'
+
+describe('/repomap argument parsing', () => {
+  test('defaults to 1024 tokens with no flags', () => {
+    const result = parseArgs('')
+    expect(result.tokens).toBe(2048)
+    expect(result.focus).toEqual([])
+    expect(result.invalidate).toBe(false)
+    expect(result.stats).toBe(false)
+  })
+
+  test('parses --tokens flag', () => {
+    const result = parseArgs('--tokens 4096')
+    expect(result.tokens).toBe(4096)
+  })
+
+  test('rejects --tokens below 256', () => {
+    const result = parseArgs('--tokens 100')
+    expect(result.tokens).toBe(2048) // falls back to default
+  })
+
+  test('rejects --tokens above 16384', () => {
+    const result = parseArgs('--tokens 20000')
+    expect(result.tokens).toBe(2048) // falls back to default
+  })
+
+  test('parses --focus flag', () => {
+    const result = parseArgs('--focus src/tools/')
+    expect(result.focus).toEqual(['src/tools/'])
+  })
+
+  test('parses multiple --focus flags', () => {
+    const result = parseArgs('--focus src/tools/ --focus src/context.ts')
+    expect(result.focus).toEqual(['src/tools/', 'src/context.ts'])
+  })
+
+  test('parses --invalidate flag', () => {
+    const result = parseArgs('--invalidate')
+    expect(result.invalidate).toBe(true)
+    expect(result.stats).toBe(false)
+  })
+
+  test('parses --stats flag', () => {
+    const result = parseArgs('--stats')
+    expect(result.stats).toBe(true)
+    expect(result.invalidate).toBe(false)
+  })
+
+  test('parses combined flags', () => {
+    const result = parseArgs('--tokens 2048 --focus src/tools/ --invalidate')
+    expect(result.tokens).toBe(2048)
+    expect(result.focus).toEqual(['src/tools/'])
+    expect(result.invalidate).toBe(true)
+  })
+})
--- a/src/commands/repomap/repomap.ts
+++ b/src/commands/repomap/repomap.ts
@@ -0,0 +1,93 @@
+import type { LocalCommandCall } from '../../types/command.js'
+import { getCwd } from '../../utils/cwd.js'
+
+/** Parse CLI-style arguments from the command string. */
+export function parseArgs(args: string): {
+  tokens: number
+  focus: string[]
+  invalidate: boolean
+  stats: boolean
+} {
+  const parts = args.trim().split(/\s+/).filter(Boolean)
+  let tokens = 2048
+  const focus: string[] = []
+  let invalidate = false
+  let stats = false
+
+  for (let i = 0; i < parts.length; i++) {
+    const part = parts[i]!
+    if (part === '--tokens' && i + 1 < parts.length) {
+      const n = parseInt(parts[i + 1]!, 10)
+      if (!isNaN(n) && n >= 256 && n <= 16384) {
+        tokens = n
+      }
+      i++
+    } else if (part === '--focus' && i + 1 < parts.length) {
+      focus.push(parts[i + 1]!)
+      i++
+    } else if (part === '--invalidate') {
+      invalidate = true
+    } else if (part === '--stats') {
+      stats = true
+    }
+  }
+
+  return { tokens, focus, invalidate, stats }
+}
+
+export const call: LocalCommandCall = async (args) => {
+  const root = getCwd()
+  const { tokens, focus, invalidate, stats } = parseArgs(args ?? '')
+
+  // Lazy import to avoid loading tree-sitter at startup
+  const {
+    buildRepoMap,
+    invalidateCache,
+    getCacheStats,
+  } = await import('../../context/repoMap/index.js')
+
+  if (stats) {
+    const cacheStats = getCacheStats(root)
+    const lines = [
+      `Repository map cache stats:`,
+      `  Cache directory: ${cacheStats.cacheDir}`,
+      `  Cache file: ${cacheStats.cacheFile ?? '(none)'}`,
+      `  Cached entries: ${cacheStats.entryCount}`,
+      `  Cache exists: ${cacheStats.exists}`,
+    ]
+    return { type: 'text', value: lines.join('\n') }
+  }
+
+  if (invalidate) {
+    invalidateCache(root)
+    const result = await buildRepoMap({
+      root,
+      maxTokens: tokens,
+      focusFiles: focus.length > 0 ? focus : undefined,
+    })
+    return {
+      type: 'text',
+      value: [
+        `Cache invalidated and rebuilt.`,
+        `Files: ${result.fileCount} ranked (${result.totalFileCount} total) | Tokens: ${result.tokenCount} | Time: ${result.buildTimeMs}ms | Cache hit: ${result.cacheHit}`,
+        '',
+        result.map,
+      ].join('\n'),
+    }
+  }
+
+  const result = await buildRepoMap({
+    root,
+    maxTokens: tokens,
+    focusFiles: focus.length > 0 ? focus : undefined,
+  })
+
+  return {
+    type: 'text',
+    value: [
+      `Repository map: ${result.fileCount} files ranked (${result.totalFileCount} total) | Tokens: ${result.tokenCount} | Time: ${result.buildTimeMs}ms | Cache hit: ${result.cacheHit}`,
+      '',
+      result.map,
+    ].join('\n'),
+  }
+}
--- a/src/components/ProviderManager.tsx
+++ b/src/components/ProviderManager.tsx
@@ -3,14 +3,12 @@ import * as React from 'react'
 import { DEFAULT_CODEX_BASE_URL } from '../services/api/providerConfig.js'
 import { Box, Text } from '../ink.js'
 import { useKeybinding } from '../keybindings/useKeybinding.js'
-import { useSetAppState } from '../state/AppState.js'
 import type { ProviderProfile } from '../utils/config.js'
 import {
  clearCodexCredentials,
  readCodexCredentialsAsync,
 } from '../utils/codexCredentials.js'
 import { isBareMode, isEnvTruthy } from '../utils/envUtils.js'
-import { getPrimaryModel, hasMultipleModels, parseModelList } from '../utils/providerModels.js'
 import {
  applySavedProfileToCurrentSession,
  buildCodexOAuthProfileEnv,
@@ -52,7 +50,6 @@ import {
 import { Pane } from './design-system/Pane.js'
 import TextInput from './TextInput.js'
 import { useCodexOAuthFlow } from './useCodexOAuthFlow.js'
-import { useSetAppState } from '../state/AppState.js'

 export type ProviderManagerResult = {
  action: 'saved' | 'cancelled'
@@ -111,8 +108,8 @@ const FORM_STEPS: Array<{
  {
    key: 'model',
    label: 'Default model',
-    placeholder: 'e.g. llama3.1:8b or glm-4.7, glm-4.7-flash',
-    helpText: 'Model name(s) to use. Separate multiple with commas; first is default.',
+    placeholder: 'e.g. llama3.1:8b',
+    helpText: 'Model name to use when this provider is active.',
  },
  {
    key: 'apiKey',
@@ -156,12 +153,7 @@ function profileSummary(profile: ProviderProfile, isActive: boolean): string {
  const keyInfo = profile.apiKey ? 'key set' : 'no key'
  const providerKind =
    profile.provider === 'anthropic' ? 'anthropic' : 'openai-compatible'
-  const models = parseModelList(profile.model)
-  const modelDisplay =
-    models.length <= 3
-      ? models.join(', ')
-      : `${models[0]}, ${models[1]} + ${models.length - 2} more`
-  return `${providerKind} · ${profile.baseUrl} · ${modelDisplay} · ${keyInfo}${activeSuffix}`
+  return `${providerKind} · ${profile.baseUrl} · ${profile.model} · ${keyInfo}${activeSuffix}`
 }

 function getGithubCredentialSourceFromEnv(
@@ -328,7 +320,6 @@ function CodexOAuthSetup({
 }

 export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
-  const setAppState = useSetAppState()
  const initialGithubCredentialSource = getGithubCredentialSourceFromEnv()
  const initialIsGithubActive = isEnvTruthy(process.env.CLAUDE_CODE_USE_GITHUB)
  const initialHasGithubCredential = initialGithubCredentialSource !== 'none'
@@ -362,7 +353,6 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
  const [cursorOffset, setCursorOffset] = React.useState(0)
  const [statusMessage, setStatusMessage] = React.useState<string | undefined>()
  const [errorMessage, setErrorMessage] = React.useState<string | undefined>()
-  const [menuFocusValue, setMenuFocusValue] = React.useState<string | undefined>()
  const [hasStoredCodexOAuthCredentials, setHasStoredCodexOAuthCredentials] =
    React.useState(false)
  const [storedCodexOAuthProfileId, setStoredCodexOAuthProfileId] =
@@ -578,48 +568,24 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
        const githubError = activateGithubProvider()
        if (githubError) {
          setErrorMessage(`Could not activate GitHub provider: ${githubError}`)
-          returnToMenu()
+          setScreen('menu')
          return
        }

-        setAppState(prev => ({
-          ...prev,
-          mainLoopModel: GITHUB_PROVIDER_DEFAULT_MODEL,
-          mainLoopModelForSession: null,
-        }))
        refreshProfiles()
-        setAppState(prev => ({
-          ...prev,
-          mainLoopModel: GITHUB_PROVIDER_DEFAULT_MODEL,
-        }))
        setStatusMessage(`Active provider: ${GITHUB_PROVIDER_LABEL}`)
-        returnToMenu()
+        setScreen('menu')
        return
      }

      const active = setActiveProviderProfile(profileId)
      if (!active) {
        setErrorMessage('Could not change active provider.')
-        returnToMenu()
+        setScreen('menu')
        return
      }

-      // Update the session model to the new provider's first model.
-      // persistActiveProviderProfileModel (called by onChangeAppState) will
-      // not overwrite the multi-model list because it checks if the model
-      // is already in the profile's comma-separated model list.
-      const newModel = getPrimaryModel(active.model)
-      setAppState(prev => ({
-        ...prev,
-        mainLoopModel: newModel,
-      }))
-
      providerLabel = active.name
-      setAppState(prev => ({
-        ...prev,
-        mainLoopModel: active.model,
-        mainLoopModelForSession: null,
-      }))
      const settingsOverrideError =
        clearStartupProviderOverrideFromUserSettings()
      const isActiveCodexOAuth = isCodexOAuthProfile(
@@ -647,21 +613,16 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
            ? `Active provider: ${active.name}. Warning: could not clear startup provider override (${settingsOverrideError}).`
            : `Active provider: ${active.name}`,
      )
-      returnToMenu()
+      setScreen('menu')
    } catch (error) {
      refreshProfiles()
      setStatusMessage(undefined)
      const detail = error instanceof Error ? error.message : String(error)
      setErrorMessage(`Could not finish activating ${providerLabel}: ${detail}`)
-      returnToMenu()
+      setScreen('menu')
    }
  }

-  function returnToMenu(): void {
-    setMenuFocusValue('done')
-    setScreen('menu')
-  }
-
  function closeWithCancelled(message: string): void {
    onDone({ action: 'cancelled', message })
  }
@@ -812,13 +773,6 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
    }

    const isActiveSavedProfile = getActiveProviderProfile()?.id === saved.id
-    if (isActiveSavedProfile) {
-      setAppState(prev => ({
-        ...prev,
-        mainLoopModel: saved.model,
-        mainLoopModelForSession: null,
-      }))
-    }
    const settingsOverrideError = isActiveSavedProfile
      ? clearStartupProviderOverrideFromUserSettings()
      : null
@@ -846,7 +800,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
    setEditingProfileId(null)
    setFormStepIndex(0)
    setErrorMessage(undefined)
-    returnToMenu()
+    setScreen('menu')
  }

  function renderOllamaSelection(): React.ReactNode {
@@ -969,7 +923,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
      return
    }

-    returnToMenu()
+    setScreen('menu')
  }

  useKeybinding('confirm:no', handleBackFromForm, {
@@ -1050,31 +1004,11 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
        label: 'LM Studio',
        description: 'Local LM Studio endpoint',
      },
-      {
-        value: 'dashscope-cn',
-        label: 'Alibaba Coding Plan (China)',
-        description: 'Alibaba DashScope China endpoint',
-      },
-      {
-        value: 'dashscope-intl',
-        label: 'Alibaba Coding Plan',
-        description: 'Alibaba DashScope International endpoint',
-      },
      {
        value: 'custom',
        label: 'Custom',
        description: 'Any OpenAI-compatible provider',
      },
-      {
-        value: 'nvidia-nim',
-        label: 'NVIDIA NIM',
-        description: 'NVIDIA NIM endpoint',
-      },
-      {
-        value: 'minimax',
-        label: 'MiniMax',
-        description: 'MiniMax API endpoint',
-      },
      ...(mode === 'first-run'
        ? [
            {
@@ -1112,7 +1046,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
              closeWithCancelled('Provider setup skipped')
              return
            }
-            returnToMenu()
+            setScreen('menu')
          }}
          visibleOptionCount={Math.min(13, options.length)}
        />
@@ -1150,7 +1084,6 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
            focus={true}
            showCursor={true}
            placeholder={`${currentStep.placeholder}${figures.ellipsis}`}
-            mask={currentStepKey === 'apiKey' ? '*' : undefined}
            columns={80}
            cursorOffset={cursorOffset}
            onChangeCursorOffset={setCursorOffset}
@@ -1313,7 +1246,6 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
            }
          }}
          onCancel={() => closeWithCancelled('Provider manager closed')}
-          defaultFocusValue={menuFocusValue}
          visibleOptionCount={options.length}
        />
      </Box>
@@ -1361,8 +1293,8 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
                description: 'Return to provider manager',
              },
            ]}
-            onChange={() => returnToMenu()}
-            onCancel={() => returnToMenu()}
+            onChange={() => setScreen('menu')}
+            onCancel={() => setScreen('menu')}
            visibleOptionCount={1}
          />
        </Box>
@@ -1377,7 +1309,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
        <Select
          options={selectOptions}
          onChange={onSelect}
-          onCancel={() => returnToMenu()}
+          onCancel={() => setScreen('menu')}
          visibleOptionCount={Math.min(10, Math.max(2, selectOptions.length))}
        />
      </Box>
@@ -1418,7 +1350,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
              setErrorMessage(
                'Codex OAuth login finished, but the provider profile could not be saved.',
              )
-              returnToMenu()
+              setScreen('menu')
              return
            }

@@ -1430,7 +1362,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
              setErrorMessage(
                'Codex OAuth login finished, but the provider could not be set as the startup provider.',
              )
-              returnToMenu()
+              setScreen('menu')
              return
            }

@@ -1464,7 +1396,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {

            setStatusMessage(message)
            setErrorMessage(undefined)
-            returnToMenu()
+            setScreen('menu')
          }}
        />
      )
@@ -1504,7 +1436,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
              refreshProfiles()
              setStatusMessage('GitHub provider deleted')
            }
-            returnToMenu()
+            setScreen('menu')
            return
          }

@@ -1539,7 +1471,7 @@ export function ProviderManager({ mode, onDone }: Props): React.ReactNode {
                : 'Provider deleted',
            )
          }
-          returnToMenu()
+          setScreen('menu')
        },
        { includeGithub: true },
      )
--- a/src/components/StartupScreen.ts
+++ b/src/components/StartupScreen.ts
@@ -117,28 +117,17 @@ function detectProvider(): { name: string; model: string; baseUrl: string; isLoc
    const baseUrl = resolvedRequest.baseUrl
    const isLocal = isLocalProviderUrl(baseUrl)
    let name = 'OpenAI'
-    if (/nvidia/i.test(baseUrl) || /nvidia/i.test(rawModel) || process.env.NVIDIA_NIM)
-      name = 'NVIDIA NIM'
-    else if (/minimax/i.test(baseUrl) || /minimax/i.test(rawModel) || process.env.MINIMAX_API_KEY)
-      name = 'MiniMax'
-    else if (resolvedRequest.transport === 'codex_responses' || baseUrl.includes('chatgpt.com/backend-api/codex'))
+    // Override to Codex when resolved endpoint is Codex
+    if (resolvedRequest.transport === 'codex_responses' || baseUrl.includes('chatgpt.com/backend-api/codex')) {
      name = 'Codex'
-    else if (/deepseek/i.test(baseUrl) || /deepseek/i.test(rawModel))
-      name = 'DeepSeek'
-    else if (/openrouter/i.test(baseUrl))
-      name = 'OpenRouter'
-    else if (/together/i.test(baseUrl))
-      name = 'Together AI'
-    else if (/groq/i.test(baseUrl))
-      name = 'Groq'
-    else if (/mistral/i.test(baseUrl) || /mistral/i.test(rawModel))
-      name = 'Mistral'
-    else if (/azure/i.test(baseUrl))
-      name = 'Azure OpenAI'
-    else if (/llama/i.test(rawModel))
-      name = 'Meta Llama'
-    else if (isLocal)
-      name = getLocalOpenAICompatibleProviderLabel(baseUrl)
+    } else if (/deepseek/i.test(baseUrl) || /deepseek/i.test(rawModel))       name = 'DeepSeek'
+    else if (/openrouter/i.test(baseUrl))                             name = 'OpenRouter'
+    else if (/together/i.test(baseUrl))                               name = 'Together AI'
+    else if (/groq/i.test(baseUrl))                                   name = 'Groq'
+    else if (/mistral/i.test(baseUrl) || /mistral/i.test(rawModel))     name = 'Mistral'
+    else if (/azure/i.test(baseUrl))                                  name = 'Azure OpenAI'
+    else if (/llama/i.test(rawModel))                                    name = 'Meta Llama'
+    else if (isLocal)                                                  name = getLocalOpenAICompatibleProviderLabel(baseUrl)
    
    // Resolve model alias to actual model name + reasoning effort
    let displayModel = resolvedRequest.resolvedModel
@@ -153,9 +142,7 @@ function detectProvider(): { name: string; model: string; baseUrl: string; isLoc
  const settings = getSettings_DEPRECATED() || {}
  const modelSetting = settings.model || process.env.ANTHROPIC_MODEL || process.env.CLAUDE_MODEL || 'claude-sonnet-4-6'
  const resolvedModel = parseUserSpecifiedModel(modelSetting)
-  const baseUrl = process.env.ANTHROPIC_BASE_URL ?? 'https://api.anthropic.com'
-  const isLocal = isLocalProviderUrl(baseUrl)
-  return { name: 'Anthropic', model: resolvedModel, baseUrl, isLocal }
+  return { name: 'Anthropic', model: resolvedModel, baseUrl: 'https://api.anthropic.com', isLocal: false }
 }

 // ─── Box drawing ──────────────────────────────────────────────────────────────
--- a/src/components/TextInput.test.tsx
+++ b/src/components/TextInput.test.tsx
@@ -6,7 +6,6 @@ import stripAnsi from 'strip-ansi'

 import { createRoot } from '../ink.js'
 import { AppStateProvider } from '../state/AppState.js'
-import { maskTextWithVisibleEdges } from '../utils/Cursor.js'
 import TextInput from './TextInput.js'
 import VimTextInput from './VimTextInput.js'

@@ -200,13 +199,6 @@ test('TextInput renders typed characters before delayed parent value commits', a
  expect(output).not.toContain('Type here...')
 })

-test('maskTextWithVisibleEdges preserves only the first and last three chars', () => {
-  expect(maskTextWithVisibleEdges('sk-secret-12345678', '*')).toBe(
-    'sk-************678',
-  )
-  expect(maskTextWithVisibleEdges('abcdef', '*')).toBe('******')
-})
-
 test('VimTextInput preserves rapid typed characters before delayed parent value commits', async () => {
  const { stdout, stdin, getOutput } = createTestStreams()
  const root = await createRoot({
--- a/src/constants/promptIdentity.test.ts
+++ b/src/constants/promptIdentity.test.ts
@@ -1,16 +1,5 @@
 import { afterEach, expect, test } from 'bun:test'

-// MACRO is replaced at build time by Bun.define but not in test mode.
-// Define it globally so tests that import modules using MACRO don't crash.
-;(globalThis as Record<string, unknown>).MACRO = {
-  VERSION: '99.0.0',
-  DISPLAY_VERSION: '0.0.0-test',
-  BUILD_TIME: new Date().toISOString(),
-  ISSUES_EXPLAINER: 'report the issue at https://github.com/anthropics/claude-code/issues',
-  PACKAGE_URL: '@gitlawb/openclaude',
-  NATIVE_PACKAGE_URL: undefined,
-}
-
 import { getSystemPrompt, DEFAULT_AGENT_PROMPT } from './prompts.js'
 import { CLI_SYSPROMPT_PREFIXES, getCLISyspromptPrefix } from './system.js'
 import { CLAUDE_CODE_GUIDE_AGENT } from '../tools/AgentTool/built-in/claudeCodeGuideAgent.js'
--- a/src/context.repoMap.test.ts
+++ b/src/context.repoMap.test.ts
@@ -0,0 +1,64 @@
+import { afterEach, describe, expect, test } from 'bun:test'
+
+afterEach(() => {
+  delete process.env.REPO_MAP
+})
+
+describe('getRepoMapContext', () => {
+  test('returns null when REPO_MAP env flag is off (default)', async () => {
+    const { getRepoMapContext } = await import('./context.js')
+    const result = await getRepoMapContext()
+    expect(result).toBeNull()
+  })
+
+  test('buildRepoMap produces valid output for context injection', async () => {
+    process.env.REPO_MAP = '1'
+    const { mkdtempSync, writeFileSync, rmSync } = await import('fs')
+    const { tmpdir } = await import('os')
+    const { join } = await import('path')
+    const { buildRepoMap } = await import('./context/repoMap/index.js')
+
+    const tempDir = mkdtempSync(join(tmpdir(), 'repomap-ctx-'))
+    try {
+      writeFileSync(
+        join(tempDir, 'main.ts'),
+        'export function main(): void { console.log("hello") }\n',
+      )
+      writeFileSync(
+        join(tempDir, 'utils.ts'),
+        'import { main } from "./main"\nexport function helper(): void { main() }\n',
+      )
+
+      const result = await buildRepoMap({
+        root: tempDir,
+        maxTokens: 1024,
+      })
+
+      // Valid map that could be injected
+      expect(result.map.length).toBeGreaterThan(0)
+      expect(result.tokenCount).toBeGreaterThan(0)
+      expect(result.tokenCount).toBeLessThanOrEqual(1024)
+      expect(typeof result.cacheHit).toBe('boolean')
+    } finally {
+      rmSync(tempDir, { recursive: true, force: true })
+      const { invalidateCache } = await import('./context/repoMap/index.js')
+      invalidateCache(tempDir)
+    }
+  })
+
+  test('getSystemContext does not include repoMap key when flag is off', async () => {
+    const { getSystemContext } = await import('./context.js')
+    const result = await getSystemContext()
+    expect('repoMap' in result).toBe(false)
+  })
+
+  test('getSystemContext includes repoMap key when REPO_MAP env flag is on', async () => {
+    process.env.REPO_MAP = '1'
+    const { getSystemContext, getRepoMapContext } = await import('./context.js')
+    getRepoMapContext.cache.clear?.()
+    getSystemContext.cache.clear?.()
+    const result = await getSystemContext()
+    expect(typeof result.repoMap).toBe('string')
+    expect(result.repoMap!.length).toBeGreaterThan(0)
+  })
+})
--- a/src/context.ts
+++ b/src/context.ts
@@ -31,6 +31,7 @@ export function setSystemPromptInjection(value: string | null): void {
  // Clear context caches immediately when injection changes
  getUserContext.cache.clear?.()
  getSystemContext.cache.clear?.()
+  getRepoMapContext.cache.clear?.()
 }

 export const getGitStatus = memoize(async (): Promise<string | null> => {
@@ -110,6 +111,35 @@ export const getGitStatus = memoize(async (): Promise<string | null> => {
  }
 })

+export const getRepoMapContext = memoize(
+  async (): Promise<string | null> => {
+    const runtimeEnabled = isEnvTruthy(process.env.REPO_MAP)
+    if (!runtimeEnabled) return null
+    if (isBareMode()) return null
+    if (isEnvTruthy(process.env.CLAUDE_CODE_REMOTE)) return null
+
+    try {
+      const startTime = Date.now()
+      logForDiagnosticsNoPII('info', 'repo_map_started')
+      const { buildRepoMap } = await import('./context/repoMap/index.js')
+      const result = await buildRepoMap({ maxTokens: 1024 })
+      logForDiagnosticsNoPII('info', 'repo_map_completed', {
+        duration_ms: Date.now() - startTime,
+        token_count: result.tokenCount,
+        file_count: result.fileCount,
+        cache_hit: result.cacheHit,
+      })
+      if (!result.map || result.map.length === 0) return null
+      return `This is a structural map of the repository, ranked by importance. Use it to understand the codebase architecture.\n\n${result.map}`
+    } catch (err) {
+      logForDiagnosticsNoPII('warn', 'repo_map_failed', {
+        error: String(err),
+      })
+      return null
+    }
+  },
+)
+
 /**
 * This context is prepended to each conversation, and cached for the duration of the conversation.
 */
@@ -127,6 +157,8 @@ export const getSystemContext = memoize(
        ? null
        : await getGitStatus()

+    const repoMap = await getRepoMapContext()
+
    // Include system prompt injection if set (for cache breaking, internal-only)
    const injection = feature('BREAK_CACHE_COMMAND')
      ? getSystemPromptInjection()
@@ -135,11 +167,13 @@ export const getSystemContext = memoize(
    logForDiagnosticsNoPII('info', 'system_context_completed', {
      duration_ms: Date.now() - startTime,
      has_git_status: gitStatus !== null,
+      has_repo_map: repoMap !== null,
      has_injection: injection !== null,
    })

    return {
      ...(gitStatus && { gitStatus }),
+      ...(repoMap && { repoMap }),
      ...(feature('BREAK_CACHE_COMMAND') && injection
        ? {
            cacheBreaker: `[CACHE_BREAKER: ${injection}]`,
--- a/src/context/repoMap/fixtures/mini-repo/fileA.ts
+++ b/src/context/repoMap/fixtures/mini-repo/fileA.ts
@@ -0,0 +1,29 @@
+// fileA — imports from fileB and fileC
+
+import { CacheLayer, buildCache } from './fileB'
+import { createStore, type StoreConfig } from './fileC'
+
+export class AppController {
+  private cache: CacheLayer
+  private config: StoreConfig
+
+  constructor(config: StoreConfig) {
+    this.cache = buildCache()
+    this.config = config
+  }
+
+  initialize(): void {
+    const store = createStore()
+    this.cache.cacheSet('primary', store)
+  }
+
+  getFromCache(key: string): unknown {
+    return this.cache.cacheGet(key)
+  }
+}
+
+export function startApp(config: StoreConfig): AppController {
+  const app = new AppController(config)
+  app.initialize()
+  return app
+}
--- a/src/context/repoMap/fixtures/mini-repo/fileB.ts
+++ b/src/context/repoMap/fixtures/mini-repo/fileB.ts
@@ -0,0 +1,23 @@
+// fileB — imports from fileC
+
+import { DataStore, createStore } from './fileC'
+
+export class CacheLayer {
+  private store: DataStore
+
+  constructor() {
+    this.store = createStore()
+  }
+
+  cacheGet(key: string): unknown | undefined {
+    return this.store.lookup(key)
+  }
+
+  cacheSet(key: string, value: unknown): void {
+    this.store.add(key, value)
+  }
+}
+
+export function buildCache(): CacheLayer {
+  return new CacheLayer()
+}
--- a/src/context/repoMap/fixtures/mini-repo/fileC.ts
+++ b/src/context/repoMap/fixtures/mini-repo/fileC.ts
@@ -0,0 +1,22 @@
+// fileC — the most imported module (imported by fileA and fileB)
+
+export class DataStore {
+  private items: Map<string, unknown> = new Map()
+
+  add(key: string, value: unknown): void {
+    this.items.set(key, value)
+  }
+
+  lookup(key: string): unknown | undefined {
+    return this.items.get(key)
+  }
+}
+
+export function createStore(): DataStore {
+  return new DataStore()
+}
+
+export interface StoreConfig {
+  maxSize: number
+  ttl: number
+}
--- a/src/context/repoMap/fixtures/mini-repo/fileD.ts
+++ b/src/context/repoMap/fixtures/mini-repo/fileD.ts
@@ -0,0 +1,9 @@
+// fileD — imports from fileA
+
+import { AppController, startApp } from './fileA'
+
+export function runApp(): void {
+  const controller: AppController = startApp({ maxSize: 100, ttl: 3600 })
+  const result = controller.getFromCache('test')
+  console.log(result)
+}
--- a/src/context/repoMap/fixtures/mini-repo/fileE.ts
+++ b/src/context/repoMap/fixtures/mini-repo/fileE.ts
@@ -0,0 +1,25 @@
+// fileE — isolated, no imports from other fixture files
+
+export interface Logger {
+  log(message: string): void
+  warn(message: string): void
+  error(message: string): void
+}
+
+export class ConsoleLogger implements Logger {
+  log(message: string): void {
+    console.log(`[LOG] ${message}`)
+  }
+
+  warn(message: string): void {
+    console.warn(`[WARN] ${message}`)
+  }
+
+  error(message: string): void {
+    console.error(`[ERROR] ${message}`)
+  }
+}
+
+export function createLogger(): Logger {
+  return new ConsoleLogger()
+}
--- a/src/context/repoMap/cache.ts
+++ b/src/context/repoMap/cache.ts
@@ -0,0 +1,139 @@
+import { createHash } from 'crypto'
+import {
+  existsSync,
+  mkdirSync,
+  readFileSync,
+  statSync,
+  writeFileSync,
+} from 'fs'
+import { homedir } from 'os'
+import { join } from 'path'
+import type { CacheData, CacheEntry, CacheStats, Tag } from './types.js'
+
+const CACHE_VERSION = 1
+const CACHE_DIR = join(homedir(), '.openclaude', 'repomap-cache')
+
+function getCacheFilePath(root: string): string {
+  const hash = createHash('sha1').update(root).digest('hex')
+  return join(CACHE_DIR, `${hash}.json`)
+}
+
+function ensureCacheDir(): void {
+  if (!existsSync(CACHE_DIR)) {
+    mkdirSync(CACHE_DIR, { recursive: true })
+  }
+}
+
+/** Load cache from disk. Returns empty cache if not found or invalid. */
+export function loadCache(root: string): CacheData {
+  const path = getCacheFilePath(root)
+  try {
+    const raw = readFileSync(path, 'utf-8')
+    const data = JSON.parse(raw) as CacheData
+    if (data.version !== CACHE_VERSION) {
+      return { version: CACHE_VERSION, entries: {} }
+    }
+    return data
+  } catch {
+    return { version: CACHE_VERSION, entries: {} }
+  }
+}
+
+/** Save cache to disk. */
+export function saveCache(root: string, cache: CacheData): void {
+  ensureCacheDir()
+  const path = getCacheFilePath(root)
+  writeFileSync(path, JSON.stringify(cache), 'utf-8')
+}
+
+/**
+ * Check if a file's cached entry is still valid based on mtime and size.
+ * Returns the cached tags if valid, null otherwise.
+ */
+export function getCachedTags(
+  cache: CacheData,
+  filePath: string,
+  root: string,
+): Tag[] | null {
+  const entry = cache.entries[filePath]
+  if (!entry) return null
+
+  try {
+    const absolutePath = join(root, filePath)
+    const stat = statSync(absolutePath)
+    if (stat.mtimeMs === entry.mtimeMs && stat.size === entry.size) {
+      return entry.tags
+    }
+  } catch {
+    // File may have been deleted
+  }
+  return null
+}
+
+/** Update the cache entry for a file. */
+export function setCachedTags(
+  cache: CacheData,
+  filePath: string,
+  root: string,
+  tags: Tag[],
+): void {
+  try {
+    const absolutePath = join(root, filePath)
+    const stat = statSync(absolutePath)
+    cache.entries[filePath] = {
+      tags,
+      mtimeMs: stat.mtimeMs,
+      size: stat.size,
+    }
+  } catch {
+    // If we can't stat, don't cache
+  }
+}
+
+/**
+ * Compute a hash of the inputs that affect the rendered map.
+ * Used to cache the final rendered output.
+ */
+export function computeMapHash(
+  files: string[],
+  maxTokens: number,
+  focusFiles: string[],
+): string {
+  const sorted = [...files].sort()
+  const input = JSON.stringify({ files: sorted, maxTokens, focusFiles: [...focusFiles].sort() })
+  return createHash('sha1').update(input).digest('hex')
+}
+
+/** Get cache statistics. */
+export function getCacheStats(root: string): CacheStats {
+  const cacheFile = getCacheFilePath(root)
+  const exists = existsSync(cacheFile)
+  let entryCount = 0
+
+  if (exists) {
+    try {
+      const data = JSON.parse(readFileSync(cacheFile, 'utf-8')) as CacheData
+      entryCount = Object.keys(data.entries).length
+    } catch {
+      // corrupted cache
+    }
+  }
+
+  return {
+    cacheDir: CACHE_DIR,
+    cacheFile: exists ? cacheFile : null,
+    entryCount,
+    exists,
+  }
+}
+
+/** Delete the cache for a repo root. */
+export function invalidateCache(root: string): void {
+  const path = getCacheFilePath(root)
+  try {
+    const { unlinkSync } = require('fs')
+    unlinkSync(path)
+  } catch {
+    // File may not exist
+  }
+}
--- a/src/context/repoMap/gitFiles.ts
+++ b/src/context/repoMap/gitFiles.ts
@@ -0,0 +1,109 @@
+import { execFile } from 'child_process'
+import { readdirSync } from 'fs'
+import { join, relative } from 'path'
+import type { SupportedLanguage } from './types.js'
+
+const SUPPORTED_EXTENSIONS: Record<string, SupportedLanguage> = {
+  '.ts': 'typescript',
+  '.tsx': 'typescript',
+  '.js': 'javascript',
+  '.jsx': 'javascript',
+  '.mjs': 'javascript',
+  '.cjs': 'javascript',
+  '.py': 'python',
+}
+
+const EXCLUDED_DIRS = new Set([
+  'node_modules',
+  'dist',
+  '.git',
+  '.hg',
+  '.svn',
+  'build',
+  'out',
+  'coverage',
+  '__pycache__',
+  '.next',
+  '.nuxt',
+  'vendor',
+  '.worktrees',
+])
+
+const EXCLUDED_FILES = new Set([
+  'bun.lock',
+  'bun.lockb',
+  'package-lock.json',
+  'yarn.lock',
+  'pnpm-lock.yaml',
+])
+
+export function getLanguageForFile(filePath: string): SupportedLanguage | null {
+  const ext = filePath.substring(filePath.lastIndexOf('.'))
+  return SUPPORTED_EXTENSIONS[ext] ?? null
+}
+
+export function isSupportedFile(filePath: string): boolean {
+  return getLanguageForFile(filePath) !== null
+}
+
+/** List files using git ls-files. Returns relative paths. */
+function gitLsFiles(root: string): Promise<string[]> {
+  return new Promise((resolve, reject) => {
+    execFile(
+      'git',
+      ['ls-files', '--cached', '--others', '--exclude-standard'],
+      { cwd: root, maxBuffer: 10 * 1024 * 1024 },
+      (error, stdout) => {
+        if (error) {
+          reject(error)
+          return
+        }
+        const files = stdout
+          .split('\n')
+          .map(f => f.trim())
+          .filter(f => f.length > 0)
+        resolve(files)
+      },
+    )
+  })
+}
+
+/** Walk directory tree manually as fallback when git is unavailable. */
+function walkDirectory(root: string, currentDir: string = root): string[] {
+  const results: string[] = []
+  let entries: ReturnType<typeof readdirSync>
+  try {
+    entries = readdirSync(currentDir, { withFileTypes: true })
+  } catch {
+    return results
+  }
+
+  for (const entry of entries) {
+    const name = entry.name
+    if (entry.isDirectory()) {
+      if (!EXCLUDED_DIRS.has(name) && !name.startsWith('.')) {
+        results.push(...walkDirectory(root, join(currentDir, name)))
+      }
+    } else if (entry.isFile()) {
+      if (!EXCLUDED_FILES.has(name)) {
+        results.push(relative(root, join(currentDir, name)))
+      }
+    }
+  }
+  return results
+}
+
+/**
+ * Enumerate all supported source files in the repo.
+ * Tries git ls-files first, falls back to manual walk.
+ */
+export async function getRepoFiles(root: string): Promise<string[]> {
+  let files: string[]
+  try {
+    files = await gitLsFiles(root)
+  } catch {
+    files = walkDirectory(root)
+  }
+
+  return files.filter(isSupportedFile)
+}
--- a/src/context/repoMap/graph.ts
+++ b/src/context/repoMap/graph.ts
@@ -0,0 +1,88 @@
+import Graph from 'graphology'
+import type { FileTags } from './types.js'
+
+// Common identifiers that should contribute less weight (high IDF penalty).
+const COMMON_NAMES = new Set([
+  'map', 'get', 'set', 'value', 'key', 'data', 'result', 'error',
+  'name', 'type', 'id', 'index', 'item', 'items', 'list', 'options',
+  'config', 'args', 'params', 'props', 'state', 'event', 'callback',
+  'handler', 'fn', 'func', 'self', 'this', 'ctx', 'context', 'req',
+  'res', 'next', 'err', 'msg', 'obj', 'arr', 'str', 'num', 'val',
+  'init', 'start', 'stop', 'run', 'main', 'test', 'setup', 'teardown',
+  'constructor', 'toString', 'valueOf', 'length', 'size', 'count',
+  'push', 'pop', 'shift', 'filter', 'reduce', 'forEach', 'find',
+  'log', 'warn', 'info', 'debug', 'trace',
+])
+
+/**
+ * Build a directed graph from file tags.
+ *
+ * Nodes are file paths. An edge from A to B means file A references
+ * a symbol defined in file B. Edge weight = refCount * idf(symbolName).
+ */
+export function buildGraph(allFileTags: FileTags[]): Graph {
+  const graph = new Graph({ multi: false, type: 'directed' })
+
+  // Build a map from symbol name → files that define it
+  const defIndex = new Map<string, Set<string>>()
+  for (const ft of allFileTags) {
+    for (const tag of ft.tags) {
+      if (tag.kind === 'def') {
+        let files = defIndex.get(tag.name)
+        if (!files) {
+          files = new Set()
+          defIndex.set(tag.name, files)
+        }
+        files.add(ft.path)
+      }
+    }
+  }
+
+  // Compute IDF: log(totalFiles / filesDefiningSymbol)
+  // Common names get an extra penalty
+  const totalFiles = allFileTags.length
+  function idf(symbolName: string): number {
+    const defFiles = defIndex.get(symbolName)
+    const docFreq = defFiles ? defFiles.size : 1
+    const rawIdf = Math.log(totalFiles / docFreq)
+    return COMMON_NAMES.has(symbolName) ? rawIdf * 0.1 : rawIdf
+  }
+
+  // Add all files as nodes
+  for (const ft of allFileTags) {
+    if (!graph.hasNode(ft.path)) {
+      graph.addNode(ft.path)
+    }
+  }
+
+  // Build edges: for each ref in a file, find where it's defined
+  for (const ft of allFileTags) {
+    // Count refs per target file
+    const edgeWeights = new Map<string, number>()
+
+    for (const tag of ft.tags) {
+      if (tag.kind !== 'ref') continue
+
+      const defFiles = defIndex.get(tag.name)
+      if (!defFiles) continue
+
+      const weight = idf(tag.name)
+      for (const defFile of defFiles) {
+        if (defFile === ft.path) continue // skip self-references
+        const current = edgeWeights.get(defFile) ?? 0
+        edgeWeights.set(defFile, current + weight)
+      }
+    }
+
+    for (const [target, weight] of edgeWeights) {
+      if (graph.hasEdge(ft.path, target)) {
+        graph.setEdgeAttribute(ft.path, target, 'weight',
+          graph.getEdgeAttribute(ft.path, target, 'weight') + weight)
+      } else {
+        graph.addEdge(ft.path, target, { weight })
+      }
+    }
+  }
+
+  return graph
+}
--- a/src/context/repoMap/index.ts
+++ b/src/context/repoMap/index.ts
@@ -0,0 +1,144 @@
+import {
+  computeMapHash,
+  getCachedTags,
+  getCacheStats as getCacheStatsImpl,
+  invalidateCache as invalidateCacheImpl,
+  loadCache,
+  saveCache,
+  setCachedTags,
+} from './cache.js'
+import { getRepoFiles } from './gitFiles.js'
+import { buildGraph } from './graph.js'
+import { rankFiles } from './pagerank.js'
+import { initParser } from './parser.js'
+import { renderMap } from './renderer.js'
+import { extractTags } from './symbolExtractor.js'
+import type { FileTags, RepoMapOptions, RepoMapResult, CacheStats } from './types.js'
+
+const DEFAULT_MAX_TOKENS = 2048
+
+/**
+ * Build a structural summary of a code repository.
+ *
+ * Walks the repo, extracts symbols via tree-sitter, builds an IDF-weighted
+ * reference graph, ranks files with PageRank, and renders a token-budgeted
+ * structural summary.
+ */
+export async function buildRepoMap(options: RepoMapOptions = {}): Promise<RepoMapResult> {
+  const startTime = Date.now()
+  const root = options.root ?? process.cwd()
+  const maxTokens = options.maxTokens ?? DEFAULT_MAX_TOKENS
+  const focusFiles = options.focusFiles ?? []
+
+  // Initialize tree-sitter
+  await initParser()
+
+  // Get files
+  const files = options.files ?? await getRepoFiles(root)
+  const totalFileCount = files.length
+
+  // Check if we have a cached rendered map
+  const mapHash = computeMapHash(files, maxTokens, focusFiles)
+  const cache = loadCache(root)
+
+  // Check if rendered map is cached (stored as a special entry)
+  const renderedCacheKey = `__rendered__${mapHash}`
+  const renderedEntry = cache.entries[renderedCacheKey]
+  if (renderedEntry && renderedEntry.tags.length === 1) {
+    const cachedResult = renderedEntry.tags[0]!
+    // The cached "tag" stores the rendered map in the signature field
+    // and metadata in name/line fields
+    try {
+      const meta = JSON.parse(cachedResult.name)
+      return {
+        map: cachedResult.signature,
+        cacheHit: true,
+        buildTimeMs: Date.now() - startTime,
+        fileCount: meta.fileCount ?? 0,
+        totalFileCount,
+        tokenCount: meta.tokenCount ?? 0,
+      }
+    } catch {
+      // Invalid cached data, continue with full build
+    }
+  }
+
+  // Extract tags for all files (using per-file cache).
+  // Separate cached hits from files needing extraction.
+  const allFileTags: FileTags[] = []
+  const uncachedFiles: string[] = []
+
+  for (const file of files) {
+    const cachedTags = getCachedTags(cache, file, root)
+    if (cachedTags) {
+      allFileTags.push({ path: file, tags: cachedTags })
+    } else {
+      uncachedFiles.push(file)
+    }
+  }
+
+  // Process uncached files in parallel batches
+  const BATCH_SIZE = 50
+  for (let i = 0; i < uncachedFiles.length; i += BATCH_SIZE) {
+    const batch = uncachedFiles.slice(i, i + BATCH_SIZE)
+    const results = await Promise.all(
+      batch.map(file => extractTags(file, root).catch(() => null))
+    )
+    for (let j = 0; j < results.length; j++) {
+      const fileTags = results[j]
+      if (fileTags) {
+        allFileTags.push(fileTags)
+        setCachedTags(cache, fileTags.path, root, fileTags.tags)
+      }
+    }
+  }
+
+  // Build graph and rank
+  const graph = buildGraph(allFileTags)
+  const ranked = rankFiles(graph, focusFiles)
+
+  // Build a lookup map
+  const fileTagsMap = new Map<string, FileTags>()
+  for (const ft of allFileTags) {
+    fileTagsMap.set(ft.path, ft)
+  }
+
+  // Render
+  const { map, tokenCount, fileCount } = renderMap(ranked, fileTagsMap, maxTokens)
+
+  // Cache the rendered result
+  cache.entries[renderedCacheKey] = {
+    tags: [{
+      kind: 'def',
+      name: JSON.stringify({ fileCount, tokenCount }),
+      line: 0,
+      signature: map,
+    }],
+    mtimeMs: Date.now(),
+    size: 0,
+  }
+
+  saveCache(root, cache)
+
+  return {
+    map,
+    cacheHit: false,
+    buildTimeMs: Date.now() - startTime,
+    fileCount,
+    totalFileCount,
+    tokenCount,
+  }
+}
+
+/** Invalidate the disk cache for a given repo root. */
+export function invalidateCache(root?: string): void {
+  invalidateCacheImpl(root ?? process.cwd())
+}
+
+/** Get cache statistics for a given repo root. */
+export function getCacheStats(root?: string): CacheStats {
+  return getCacheStatsImpl(root ?? process.cwd())
+}
+
+// Re-export types for convenience
+export type { RepoMapOptions, RepoMapResult, CacheStats } from './types.js'
--- a/src/context/repoMap/pagerank.ts
+++ b/src/context/repoMap/pagerank.ts
@@ -0,0 +1,57 @@
+import type Graph from 'graphology'
+import pagerank from 'graphology-pagerank'
+
+export interface RankedFile {
+  path: string
+  score: number
+}
+
+/**
+ * Run PageRank on the file reference graph.
+ *
+ * focusFiles get a 100x boost in the personalization vector so they
+ * and their neighbors rank higher.
+ *
+ * Returns files sorted by score descending.
+ */
+export function rankFiles(
+  graph: Graph,
+  focusFiles: string[] = [],
+): RankedFile[] {
+  if (graph.order === 0) return []
+
+  const hasPersonalization = focusFiles.length > 0
+
+  // graphology-pagerank accepts getEdgeWeight option
+  const scores: Record<string, number> = pagerank(graph, {
+    alpha: 0.85,
+    maxIterations: 100,
+    tolerance: 1e-6,
+    getEdgeWeight: 'weight',
+  })
+
+  // Apply focus boost post-hoc if focus files are specified
+  if (hasPersonalization) {
+    for (const file of focusFiles) {
+      if (scores[file] !== undefined) {
+        scores[file] *= 100
+      }
+    }
+
+    // Also boost direct neighbors of focus files
+    for (const file of focusFiles) {
+      if (!graph.hasNode(file)) continue
+      graph.forEachNeighbor(file, (neighbor) => {
+        if (scores[neighbor] !== undefined) {
+          scores[neighbor] *= 10
+        }
+      })
+    }
+  }
+
+  const ranked: RankedFile[] = Object.entries(scores)
+    .map(([path, score]) => ({ path, score }))
+    .sort((a, b) => b.score - a.score)
+
+  return ranked
+}
--- a/src/context/repoMap/parser.ts
+++ b/src/context/repoMap/parser.ts
@@ -0,0 +1,166 @@
+import { existsSync, readFileSync } from 'fs'
+import { join, resolve } from 'path'
+import { fileURLToPath } from 'url'
+import type { SupportedLanguage } from './types.js'
+
+// Resolve project root in both source and bundled modes.
+// In source (bun test/dev): import.meta.url is src/context/repoMap/parser.ts → go up 4 levels
+// In bundle (node dist/cli.mjs): import.meta.url is dist/cli.mjs → go up 2 levels
+const __filename = fileURLToPath(import.meta.url)
+const __projectRoot = join(
+  __filename,
+  process.env.NODE_ENV === 'test' ? '../../../../' : '../../',
+)
+
+// web-tree-sitter types
+type TreeSitterParser = {
+  parse(input: string): { rootNode: unknown }
+  setLanguage(lang: unknown): void
+  delete(): void
+}
+
+type TreeSitterLanguage = {
+  query(source: string): unknown
+}
+
+// The actual module exports { Parser, Language } as named exports
+let ParserClass: (new () => TreeSitterParser) & {
+  init(opts?: { locateFile?: (file: string) => string }): Promise<void>
+} | null = null
+let LanguageLoader: {
+  load(path: string | Uint8Array): Promise<TreeSitterLanguage>
+} | null = null
+
+let initialized = false
+const languageCache = new Map<SupportedLanguage, TreeSitterLanguage>()
+const queryCache = new Map<SupportedLanguage, string>()
+
+/** Resolve the path to the tree-sitter WASM file. */
+function getTreeSitterWasmPath(): string {
+  // Try require.resolve first (works in source mode with node_modules)
+  try {
+    const webTsDir = resolve(
+      require.resolve('web-tree-sitter/package.json'),
+      '..',
+    )
+    return join(webTsDir, 'tree-sitter.wasm')
+  } catch {
+    // Fallback: relative to project root
+    return join(__projectRoot, 'node_modules', 'web-tree-sitter', 'tree-sitter.wasm')
+  }
+}
+
+/** Resolve the path to a language WASM grammar file. */
+function getLanguageWasmPath(language: SupportedLanguage): string {
+  const wasmName = language === 'typescript' ? 'tree-sitter-typescript' :
+    language === 'javascript' ? 'tree-sitter-javascript' :
+      `tree-sitter-${language}`
+
+  try {
+    const wasmDir = resolve(
+      require.resolve('tree-sitter-wasms/package.json'),
+      '..',
+      'out',
+    )
+    return join(wasmDir, `${wasmName}.wasm`)
+  } catch {
+    return join(__projectRoot, 'node_modules', 'tree-sitter-wasms', 'out', `${wasmName}.wasm`)
+  }
+}
+
+/** Resolve the path to a tag query .scm file for the given language. */
+function getQueryPath(language: SupportedLanguage): string {
+  // Try source location first (works in both source and when queries are alongside the bundle)
+  const sourcePath = join(__projectRoot, 'src', 'context', 'repoMap', 'queries', `${language}-tags.scm`)
+  if (existsSync(sourcePath)) {
+    return sourcePath
+  }
+  // Fallback: relative to this file (source mode)
+  return join(fileURLToPath(import.meta.url), '..', 'queries', `${language}-tags.scm`)
+}
+
+/** Initialize the tree-sitter WASM module. */
+export async function initParser(): Promise<void> {
+  if (initialized) return
+
+  try {
+    const mod = await import('web-tree-sitter')
+    ParserClass = mod.Parser as typeof ParserClass
+    LanguageLoader = mod.Language as typeof LanguageLoader
+
+    const wasmPath = getTreeSitterWasmPath()
+    await ParserClass!.init({
+      locateFile: () => wasmPath,
+    })
+    initialized = true
+  } catch (err) {
+    // eslint-disable-next-line no-console
+    console.error('[repoMap] Failed to initialize tree-sitter:', err)
+    throw err
+  }
+}
+
+/** Load a language grammar. Cached after first load. */
+export async function loadLanguage(language: SupportedLanguage): Promise<TreeSitterLanguage | null> {
+  if (languageCache.has(language)) {
+    return languageCache.get(language)!
+  }
+
+  if (!initialized) {
+    await initParser()
+  }
+
+  try {
+    const wasmPath = getLanguageWasmPath(language)
+    const lang = await LanguageLoader!.load(wasmPath)
+    languageCache.set(language, lang)
+    return lang
+  } catch (err) {
+    // eslint-disable-next-line no-console
+    console.error(`[repoMap] Failed to load ${language} grammar:`, err)
+    return null
+  }
+}
+
+/** Load the tag query for a language. Cached after first load. */
+export function loadQuery(language: SupportedLanguage): string | null {
+  if (queryCache.has(language)) {
+    return queryCache.get(language)!
+  }
+
+  try {
+    const queryPath = getQueryPath(language)
+    const content = readFileSync(queryPath, 'utf-8')
+    queryCache.set(language, content)
+    return content
+  } catch {
+    return null
+  }
+}
+
+/** Create a new parser instance with the given language set. */
+export async function createParser(language: SupportedLanguage): Promise<TreeSitterParser | null> {
+  if (!initialized) {
+    await initParser()
+  }
+
+  const lang = await loadLanguage(language)
+  if (!lang) return null
+
+  try {
+    const parser = new ParserClass!()
+    parser.setLanguage(lang)
+    return parser
+  } catch {
+    return null
+  }
+}
+
+/** Clear all caches (useful for testing). */
+export function clearParserCaches(): void {
+  languageCache.clear()
+  queryCache.clear()
+  initialized = false
+  ParserClass = null
+  LanguageLoader = null
+}
--- a/src/context/repoMap/queries/javascript-tags.scm
+++ b/src/context/repoMap/queries/javascript-tags.scm
@@ -0,0 +1,92 @@
+; Source: https://github.com/Aider-AI/aider/blob/main/aider/queries/tree-sitter-languages/javascript-tags.scm
+; License: MIT (Apache-2.0 dual) — see https://github.com/Aider-AI/aider/blob/main/LICENSE
+; Copied for use in openclaude's repo-map feature.
+
+(
+  (comment)* @doc
+  .
+  (method_definition
+    name: (property_identifier) @name.definition.method) @definition.method
+  (#not-eq? @name.definition.method "constructor")
+  (#strip! @doc "^[\\s\\*/]+|^[\\s\\*/]$")
+  (#select-adjacent! @doc @definition.method)
+)
+
+(
+  (comment)* @doc
+  .
+  [
+    (class
+      name: (_) @name.definition.class)
+    (class_declaration
+      name: (_) @name.definition.class)
+  ] @definition.class
+  (#strip! @doc "^[\\s\\*/]+|^[\\s\\*/]$")
+  (#select-adjacent! @doc @definition.class)
+)
+
+(
+  (comment)* @doc
+  .
+  [
+    (function
+      name: (identifier) @name.definition.function)
+    (function_declaration
+      name: (identifier) @name.definition.function)
+    (generator_function
+      name: (identifier) @name.definition.function)
+    (generator_function_declaration
+      name: (identifier) @name.definition.function)
+  ] @definition.function
+  (#strip! @doc "^[\\s\\*/]+|^[\\s\\*/]$")
+  (#select-adjacent! @doc @definition.function)
+)
+
+(
+  (comment)* @doc
+  .
+  (lexical_declaration
+    (variable_declarator
+      name: (identifier) @name.definition.function
+      value: [(arrow_function) (function)]) @definition.function)
+  (#strip! @doc "^[\\s\\*/]+|^[\\s\\*/]$")
+  (#select-adjacent! @doc @definition.function)
+)
+
+(
+  (comment)* @doc
+  .
+  (variable_declaration
+    (variable_declarator
+      name: (identifier) @name.definition.function
+      value: [(arrow_function) (function)]) @definition.function)
+  (#strip! @doc "^[\\s\\*/]+|^[\\s\\*/]$")
+  (#select-adjacent! @doc @definition.function)
+)
+
+(assignment_expression
+  left: [
+    (identifier) @name.definition.function
+    (member_expression
+      property: (property_identifier) @name.definition.function)
+  ]
+  right: [(arrow_function) (function)]
+) @definition.function
+
+(pair
+  key: (property_identifier) @name.definition.function
+  value: [(arrow_function) (function)]) @definition.function
+
+(
+  (call_expression
+    function: (identifier) @name.reference.call) @reference.call
+  (#not-match? @name.reference.call "^(require)$")
+)
+
+(call_expression
+  function: (member_expression
+    property: (property_identifier) @name.reference.call)
+  arguments: (_) @reference.call)
+
+(new_expression
+  constructor: (_) @name.reference.class) @reference.class
--- a/src/context/repoMap/queries/python-tags.scm
+++ b/src/context/repoMap/queries/python-tags.scm
@@ -0,0 +1,16 @@
+; Source: https://github.com/Aider-AI/aider/blob/main/aider/queries/tree-sitter-languages/python-tags.scm
+; License: MIT (Apache-2.0 dual) — see https://github.com/Aider-AI/aider/blob/main/LICENSE
+; Copied for use in openclaude's repo-map feature.
+
+(class_definition
+  name: (identifier) @name.definition.class) @definition.class
+
+(function_definition
+  name: (identifier) @name.definition.function) @definition.function
+
+(call
+  function: [
+      (identifier) @name.reference.call
+      (attribute
+        attribute: (identifier) @name.reference.call)
+  ]) @reference.call
--- a/src/context/repoMap/queries/typescript-tags.scm
+++ b/src/context/repoMap/queries/typescript-tags.scm
@@ -0,0 +1,45 @@
+; Source: https://github.com/Aider-AI/aider/blob/main/aider/queries/tree-sitter-languages/typescript-tags.scm
+; License: MIT (Apache-2.0 dual) — see https://github.com/Aider-AI/aider/blob/main/LICENSE
+; Copied for use in openclaude's repo-map feature.
+
+(function_signature
+  name: (identifier) @name.definition.function) @definition.function
+
+(method_signature
+  name: (property_identifier) @name.definition.method) @definition.method
+
+(abstract_method_signature
+  name: (property_identifier) @name.definition.method) @definition.method
+
+(abstract_class_declaration
+  name: (type_identifier) @name.definition.class) @definition.class
+
+(module
+  name: (identifier) @name.definition.module) @definition.module
+
+(interface_declaration
+  name: (type_identifier) @name.definition.interface) @definition.interface
+
+(type_annotation
+  (type_identifier) @name.reference.type) @reference.type
+
+(new_expression
+  constructor: (identifier) @name.reference.class) @reference.class
+
+(function_declaration
+  name: (identifier) @name.definition.function) @definition.function
+
+(method_definition
+  name: (property_identifier) @name.definition.method) @definition.method
+
+(class_declaration
+  name: (type_identifier) @name.definition.class) @definition.class
+
+(interface_declaration
+  name: (type_identifier) @name.definition.class) @definition.class
+
+(type_alias_declaration
+  name: (type_identifier) @name.definition.type) @definition.type
+
+(enum_declaration
+  name: (identifier) @name.definition.enum) @definition.enum
--- a/src/context/repoMap/renderer.ts
+++ b/src/context/repoMap/renderer.ts
@@ -0,0 +1,72 @@
+import type { FileTags, Tag } from './types.js'
+import type { RankedFile } from './pagerank.js'
+import { countTokens } from './tokenize.js'
+
+/**
+ * Render a token-budgeted repo map from ranked files and their tags.
+ *
+ * Format per file:
+ *   path/to/file.ts:
+ *   ⋮
+ *     signature line for def 1
+ *   ⋮
+ *     signature line for def 2
+ *   ⋮
+ *
+ * Files that don't fit within the budget are dropped entirely.
+ */
+export function renderMap(
+  rankedFiles: RankedFile[],
+  fileTagsMap: Map<string, FileTags>,
+  maxTokens: number,
+): { map: string; tokenCount: number; fileCount: number } {
+  const sections: string[] = []
+  let currentTokens = 0
+  let fileCount = 0
+
+  for (const { path } of rankedFiles) {
+    const ft = fileTagsMap.get(path)
+    if (!ft) continue
+
+    // Only include definitions in the rendered output
+    const defs = ft.tags
+      .filter(t => t.kind === 'def')
+      .sort((a, b) => a.line - b.line)
+
+    if (defs.length === 0) continue
+
+    const section = renderFileSection(path, defs)
+    const sectionTokens = countTokens(section)
+
+    // Would this section bust the budget?
+    if (currentTokens + sectionTokens > maxTokens) {
+      // Don't include partial files — drop entirely
+      break
+    }
+
+    sections.push(section)
+    currentTokens += sectionTokens
+    fileCount++
+  }
+
+  const map = sections.join('\n')
+  return { map, tokenCount: currentTokens, fileCount }
+}
+
+function renderFileSection(path: string, defs: Tag[]): string {
+  const lines: string[] = [`${path}:`]
+  let lastLine = 0
+
+  for (const def of defs) {
+    // Add elision marker if there's a gap
+    if (def.line > lastLine + 1) {
+      lines.push('⋮')
+    }
+    lines.push(`  ${def.signature}`)
+    lastLine = def.line
+  }
+
+  // Trailing elision marker
+  lines.push('⋮')
+  return lines.join('\n')
+}
--- a/src/context/repoMap/repoMap.test.ts
+++ b/src/context/repoMap/repoMap.test.ts
@@ -0,0 +1,275 @@
+import { afterEach, beforeAll, describe, expect, test } from 'bun:test'
+import { cpSync, mkdtempSync, rmSync, utimesSync, writeFileSync } from 'fs'
+import { tmpdir } from 'os'
+import { join } from 'path'
+import { invalidateCache, buildRepoMap } from './index.js'
+import { extractTags } from './symbolExtractor.js'
+import { buildGraph } from './graph.js'
+import { initParser } from './parser.js'
+import { countTokens } from './tokenize.js'
+
+const FIXTURE_ROOT = join(import.meta.dir, '__fixtures__', 'mini-repo')
+const FIXTURE_FILES = ['fileA.ts', 'fileB.ts', 'fileC.ts', 'fileD.ts', 'fileE.ts']
+
+beforeAll(async () => {
+  await initParser()
+})
+
+// Clean up cache between tests to avoid cross-test interference
+afterEach(() => {
+  invalidateCache(FIXTURE_ROOT)
+})
+
+describe('symbol extraction', () => {
+  test('extracts function and class defs from a TypeScript file', async () => {
+    const result = await extractTags('fileC.ts', FIXTURE_ROOT)
+    expect(result).not.toBeNull()
+
+    const defs = result!.tags.filter(t => t.kind === 'def')
+    const defNames = defs.map(t => t.name)
+
+    expect(defNames).toContain('DataStore')
+    expect(defNames).toContain('createStore')
+    expect(defNames).toContain('StoreConfig')
+
+    // All defs should have kind='def'
+    for (const d of defs) {
+      expect(d.kind).toBe('def')
+    }
+  })
+
+  test('extracts references to imported symbols', async () => {
+    const result = await extractTags('fileA.ts', FIXTURE_ROOT)
+    expect(result).not.toBeNull()
+
+    const refs = result!.tags.filter(t => t.kind === 'ref')
+    const refNames = refs.map(t => t.name)
+
+    // fileA imports CacheLayer from fileB and StoreConfig from fileC
+    expect(refNames).toContain('CacheLayer')
+    expect(refNames).toContain('StoreConfig')
+  })
+})
+
+describe('graph', () => {
+  test('builds edges between files that reference each other\'s symbols', async () => {
+    const allTags = []
+    for (const f of FIXTURE_FILES) {
+      const tags = await extractTags(f, FIXTURE_ROOT)
+      if (tags) allTags.push(tags)
+    }
+
+    const graph = buildGraph(allTags)
+
+    // fileA imports from fileB (references CacheLayer defined in fileB)
+    expect(graph.hasEdge('fileA.ts', 'fileB.ts')).toBe(true)
+
+    // fileA imports from fileC (references StoreConfig, DataStore defined in fileC)
+    expect(graph.hasEdge('fileA.ts', 'fileC.ts')).toBe(true)
+
+    // fileB imports from fileC (references DataStore defined in fileC)
+    expect(graph.hasEdge('fileB.ts', 'fileC.ts')).toBe(true)
+
+    // fileD imports from fileA
+    expect(graph.hasEdge('fileD.ts', 'fileA.ts')).toBe(true)
+
+    // fileE is isolated — no edges to/from it
+    expect(graph.degree('fileE.ts')).toBe(0)
+  })
+})
+
+describe('pagerank', () => {
+  test('ranks the most-imported file highest', async () => {
+    const result = await buildRepoMap({
+      root: FIXTURE_ROOT,
+      maxTokens: 2048,
+      files: FIXTURE_FILES,
+    })
+
+    // The map starts with the highest-ranked file
+    const firstFile = result.map.split('\n')[0]
+    expect(firstFile).toBe('fileC.ts:')
+
+    // fileE should be ranked lowest (or near last)
+    const lines = result.map.split('\n')
+    const filePositions = FIXTURE_FILES.map(f => {
+      const idx = lines.findIndex(l => l === `${f}:`)
+      return { file: f, position: idx }
+    }).filter(x => x.position >= 0)
+    .sort((a, b) => a.position - b.position)
+
+    // fileC should be first
+    expect(filePositions[0]!.file).toBe('fileC.ts')
+
+    // fileE should be last (or among the last)
+    const lastFile = filePositions[filePositions.length - 1]!.file
+    expect(['fileD.ts', 'fileE.ts']).toContain(lastFile)
+  })
+})
+
+describe('renderer', () => {
+  test('respects the token budget within 5%', async () => {
+    const maxTokens = 500
+    const result = await buildRepoMap({
+      root: FIXTURE_ROOT,
+      maxTokens,
+      files: FIXTURE_FILES,
+    })
+
+    const actualTokens = countTokens(result.map)
+    expect(actualTokens).toBeLessThanOrEqual(maxTokens * 1.05)
+    expect(result.tokenCount).toBeLessThanOrEqual(maxTokens * 1.05)
+  })
+
+  test('drops files that don\'t fit rather than listing their names', async () => {
+    // Very tight budget — should only fit 1-2 files
+    const result = await buildRepoMap({
+      root: FIXTURE_ROOT,
+      maxTokens: 100,
+      files: FIXTURE_FILES,
+    })
+
+    // Count how many files appear as headers in the output
+    const fileHeaders = result.map.split('\n').filter(l => l.endsWith(':') && !l.startsWith(' '))
+
+    // Every file header in the output should have its signatures listed
+    for (const header of fileHeaders) {
+      // The file must have at least one signature line after it
+      const headerIdx = result.map.indexOf(header)
+      const afterHeader = result.map.slice(headerIdx + header.length)
+      // Should have content (signatures), not just the filename
+      expect(afterHeader.trim().length).toBeGreaterThan(0)
+    }
+
+    // Should have fewer files than total
+    expect(fileHeaders.length).toBeLessThan(FIXTURE_FILES.length)
+  })
+})
+
+describe('cache', () => {
+  test('second build of unchanged fixture uses the cache', async () => {
+    // First build (cold)
+    const result1 = await buildRepoMap({
+      root: FIXTURE_ROOT,
+      maxTokens: 2048,
+      files: FIXTURE_FILES,
+    })
+    expect(result1.cacheHit).toBe(false)
+
+    // Second build (warm)
+    const result2 = await buildRepoMap({
+      root: FIXTURE_ROOT,
+      maxTokens: 2048,
+      files: FIXTURE_FILES,
+    })
+    expect(result2.cacheHit).toBe(true)
+    expect(result2.buildTimeMs).toBeLessThan(result1.buildTimeMs)
+
+    // Output should be identical
+    expect(result2.map).toBe(result1.map)
+  })
+
+  test('modifying a file invalidates only that file', async () => {
+    // Create a temp copy of the fixture
+    const tempDir = mkdtempSync(join(tmpdir(), 'repomap-test-'))
+    try {
+      for (const f of FIXTURE_FILES) {
+        cpSync(join(FIXTURE_ROOT, f), join(tempDir, f))
+      }
+
+      // First build
+      const result1 = await buildRepoMap({
+        root: tempDir,
+        maxTokens: 2048,
+        files: FIXTURE_FILES,
+      })
+      expect(result1.cacheHit).toBe(false)
+
+      // Touch one file to change its mtime
+      const targetFile = join(tempDir, 'fileE.ts')
+      const now = new Date()
+      utimesSync(targetFile, now, now)
+
+      // Second build — rendered cache should be invalidated because file list hash
+      // includes the files and the rendered map hash changes with different mtimes
+      // for the per-file cache check
+      invalidateCache(tempDir)
+      const result2 = await buildRepoMap({
+        root: tempDir,
+        maxTokens: 2048,
+        files: FIXTURE_FILES,
+      })
+      // The per-file cache for fileE should miss (mtime changed),
+      // but other files should still hit the per-file cache
+      expect(result2.cacheHit).toBe(false)
+
+      // Output should still be valid
+      expect(result2.map.length).toBeGreaterThan(0)
+      expect(result2.fileCount).toBe(result1.fileCount)
+    } finally {
+      rmSync(tempDir, { recursive: true, force: true })
+      invalidateCache(tempDir)
+    }
+  })
+})
+
+describe('gitFiles', () => {
+  test('falls back gracefully when not in a git repo', async () => {
+    // Create a temp directory with source files but NO .git
+    const tempDir = mkdtempSync(join(tmpdir(), 'repomap-nogit-'))
+    try {
+      writeFileSync(
+        join(tempDir, 'hello.ts'),
+        'export function hello(): string { return "world" }\n',
+      )
+      writeFileSync(
+        join(tempDir, 'utils.ts'),
+        'export function add(a: number, b: number): number { return a + b }\n',
+      )
+
+      const result = await buildRepoMap({
+        root: tempDir,
+        maxTokens: 1024,
+      })
+
+      // Should succeed without throwing
+      expect(result.map.length).toBeGreaterThan(0)
+      expect(result.totalFileCount).toBeGreaterThan(0)
+    } finally {
+      rmSync(tempDir, { recursive: true, force: true })
+      invalidateCache(tempDir)
+    }
+  })
+})
+
+describe('error handling', () => {
+  test('no crash on malformed source file', async () => {
+    const tempDir = mkdtempSync(join(tmpdir(), 'repomap-malformed-'))
+    try {
+      // Valid file
+      writeFileSync(
+        join(tempDir, 'good.ts'),
+        'export function good(): number { return 1 }\n',
+      )
+      // Malformed file — severe syntax errors
+      writeFileSync(
+        join(tempDir, 'bad.ts'),
+        '}{}{}{export classclass [[[ function ,,, @@@ ###\n',
+      )
+
+      const result = await buildRepoMap({
+        root: tempDir,
+        maxTokens: 1024,
+        files: ['good.ts', 'bad.ts'],
+      })
+
+      // Should complete successfully
+      expect(result.map.length).toBeGreaterThan(0)
+      // The good file should be in the output
+      expect(result.map).toContain('good.ts')
+    } finally {
+      rmSync(tempDir, { recursive: true, force: true })
+      invalidateCache(tempDir)
+    }
+  })
+})
--- a/src/context/repoMap/symbolExtractor.ts
+++ b/src/context/repoMap/symbolExtractor.ts
@@ -0,0 +1,108 @@
+import { readFileSync } from 'fs'
+import { join } from 'path'
+import { getLanguageForFile } from './gitFiles.js'
+import { createParser, loadLanguage, loadQuery } from './parser.js'
+import type { FileTags, Tag } from './types.js'
+
+/**
+ * Extract definition and reference tags from a single source file.
+ * Returns null if the file can't be parsed (unsupported language, parse error, etc).
+ */
+export async function extractTags(
+  filePath: string,
+  root: string,
+): Promise<FileTags | null> {
+  const language = getLanguageForFile(filePath)
+  if (!language) return null
+
+  const absolutePath = join(root, filePath)
+  let source: string
+  try {
+    source = readFileSync(absolutePath, 'utf-8')
+  } catch {
+    return null
+  }
+
+  const lines = source.split('\n')
+
+  const parser = await createParser(language)
+  if (!parser) return null
+
+  const querySource = loadQuery(language)
+  if (!querySource) {
+    parser.delete()
+    return null
+  }
+
+  try {
+    const tree = parser.parse(source) as {
+      rootNode: unknown
+    }
+
+    const lang = await loadLanguage(language)
+    if (!lang) {
+      parser.delete()
+      return null
+    }
+
+    // Use the non-deprecated Query constructor
+    const { Query } = await import('web-tree-sitter')
+    const query = new Query(lang, querySource) as {
+      matches(rootNode: unknown): Array<{
+        pattern: number
+        captures: Array<{
+          name: string
+          node: {
+            text: string
+            startPosition: { row: number; column: number }
+            endPosition: { row: number; column: number }
+          }
+        }>
+      }>
+    }
+
+    const matches = query.matches(tree.rootNode)
+    const tags: Tag[] = []
+    const seen = new Set<string>() // dedup by kind+name+line
+
+    for (const match of matches) {
+      let name: string | null = null
+      let kind: 'def' | 'ref' | null = null
+      let subKind: string | undefined
+      let lineRow = 0
+
+      for (const capture of match.captures) {
+        const captureName = capture.name
+
+        // Name captures: name.definition.X or name.reference.X
+        if (captureName.startsWith('name.definition.')) {
+          name = capture.node.text
+          kind = 'def'
+          subKind = captureName.slice('name.definition.'.length)
+          lineRow = capture.node.startPosition.row
+        } else if (captureName.startsWith('name.reference.')) {
+          name = capture.node.text
+          kind = 'ref'
+          subKind = captureName.slice('name.reference.'.length)
+          lineRow = capture.node.startPosition.row
+        }
+      }
+
+      if (name && kind) {
+        const key = `${kind}:${name}:${lineRow}`
+        if (!seen.has(key)) {
+          seen.add(key)
+          const line = lineRow + 1 // convert 0-based to 1-based
+          const signature = lines[lineRow]?.trimEnd() ?? ''
+          tags.push({ kind, name, line, signature, subKind })
+        }
+      }
+    }
+
+    parser.delete()
+    return { path: filePath, tags }
+  } catch {
+    parser.delete()
+    return null
+  }
+}
--- a/src/context/repoMap/tokenize.ts
+++ b/src/context/repoMap/tokenize.ts
@@ -0,0 +1,15 @@
+import { getEncoding, type Tiktoken } from 'js-tiktoken'
+
+let encoder: Tiktoken | null = null
+
+function getEncoder() {
+  if (!encoder) {
+    encoder = getEncoding('cl100k_base')
+  }
+  return encoder
+}
+
+/** Count the number of tokens in a string using cl100k_base encoding. */
+export function countTokens(text: string): number {
+  return getEncoder().encode(text).length
+}
--- a/src/context/repoMap/types.ts
+++ b/src/context/repoMap/types.ts
@@ -0,0 +1,65 @@
+export interface Tag {
+  /** 'def' for definitions, 'ref' for references */
+  kind: 'def' | 'ref'
+  /** Symbol name (e.g. function name, class name) */
+  name: string
+  /** 1-based line number in the source file */
+  line: number
+  /** The full line of source code at this position (used as signature for defs) */
+  signature: string
+  /** Sub-kind from the query (e.g. 'function', 'class', 'method', 'type') */
+  subKind?: string
+}
+
+export interface FileTags {
+  /** Relative path from the repo root */
+  path: string
+  /** All tags extracted from this file */
+  tags: Tag[]
+}
+
+export interface RepoMapOptions {
+  /** Root directory of the repo (defaults to cwd) */
+  root?: string
+  /** Maximum token budget for the rendered map */
+  maxTokens?: number
+  /** Files to boost in PageRank (relative paths) */
+  focusFiles?: string[]
+  /** Override the list of files to process (relative paths) */
+  files?: string[]
+}
+
+export interface RepoMapResult {
+  /** The rendered repo map string */
+  map: string
+  /** Whether the result came from cache */
+  cacheHit: boolean
+  /** Time in milliseconds to build the map */
+  buildTimeMs: number
+  /** Number of files included in the rendered map */
+  fileCount: number
+  /** Total number of files processed */
+  totalFileCount: number
+  /** Actual token count of the rendered map */
+  tokenCount: number
+}
+
+export interface CacheEntry {
+  tags: Tag[]
+  mtimeMs: number
+  size: number
+}
+
+export interface CacheData {
+  version: number
+  entries: Record<string, CacheEntry>
+}
+
+export interface CacheStats {
+  cacheDir: string
+  cacheFile: string | null
+  entryCount: number
+  exists: boolean
+}
+
+export type SupportedLanguage = 'typescript' | 'javascript' | 'python'
--- a/src/entrypoints/cli.tsx
+++ b/src/entrypoints/cli.tsx
@@ -5,7 +5,7 @@ import {
 } from '../utils/providerProfile.js'
 import {
  getProviderValidationError,
-  validateProviderEnvForStartupOrExit,
+  validateProviderEnvOrExit,
 } from '../utils/providerValidation.js'

 // OpenClaude: polyfill globalThis.File for Node < 20.
@@ -132,7 +132,7 @@ async function main(): Promise<void> {
    hydrateGithubModelsTokenFromSecureStorage()
  }

-  await validateProviderEnvForStartupOrExit()
+  await validateProviderEnvOrExit()

  // Print the gradient startup screen before the Ink UI loads
  const { printStartupScreen } = await import('../components/StartupScreen.js')
--- a/src/entrypoints/mcp.test.ts
+++ b/src/entrypoints/mcp.test.ts
@@ -1,75 +0,0 @@
-import { describe, it, expect, mock } from 'bun:test'
-import { getCombinedTools, loadReexposedMcpTools } from './mcp.js'
-import type { Tool as InternalTool } from '../Tool.js'
-import type { MCPServerConnection } from '../services/mcp/types.js'
-import type { Tool } from '@modelcontextprotocol/sdk/types.js'
-
-// Mock the MCP client service to control the tools and connections returned
-const mockGetMcpToolsCommandsAndResources = mock(async (onConnectionAttempt: any) => {})
-mock.module('../services/mcp/client.js', () => ({
-  getMcpToolsCommandsAndResources: mockGetMcpToolsCommandsAndResources
-}))
-
-describe('getCombinedTools', () => {
-  it('deduplicates builtins when mcpTools have the same name, prioritizing mcpTools', () => {
-    const builtinBash = { name: 'Bash', isMcp: false } as unknown as InternalTool
-    const builtinRead = { name: 'Read', isMcp: false } as unknown as InternalTool
-    const mcpBash = { name: 'Bash', isMcp: true } as unknown as InternalTool
-
-    const builtins = [builtinBash, builtinRead]
-    const mcpTools = [mcpBash]
-
-    const result = getCombinedTools(builtins, mcpTools)
-
-    expect(result).toHaveLength(2)
-    expect(result[0]).toBe(mcpBash)
-    expect(result[1]).toBe(builtinRead)
-  })
-})
-
-describe('loadReexposedMcpTools', () => {
-  it('loads tools and clients regardless of connection state (including needs-auth)', async () => {
-    // Setup the mock to simulate yielding a needs-auth server and a connected server
-    mockGetMcpToolsCommandsAndResources.mockImplementation(async (onConnectionAttempt) => {
-      const needsAuthClient = {
-        name: 'auth-server',
-        type: 'needs-auth',
-        config: {}
-      } as MCPServerConnection
-
-      const authTool = {
-        name: 'mcp__auth-server__authenticate',
-        isMcp: true
-      } as unknown as InternalTool
-
-      const connectedClient = {
-        name: 'connected-server',
-        type: 'connected',
-        config: {},
-        client: {}
-      } as MCPServerConnection
-
-      const connectedTool = {
-        name: 'mcp__connected-server__do_thing',
-        isMcp: true
-      } as unknown as InternalTool
-
-      // Simulate the callback behavior
-      onConnectionAttempt({ client: needsAuthClient, tools: [authTool], commands: [] })
-      onConnectionAttempt({ client: connectedClient, tools: [connectedTool], commands: [] })
-    })
-
-    const { mcpClients, mcpTools } = await loadReexposedMcpTools()
-
-    expect(mcpClients).toHaveLength(2)
-    expect(mcpClients[0].type).toBe('needs-auth')
-    expect(mcpClients[1].type).toBe('connected')
-
-    expect(mcpTools).toHaveLength(2)
-    expect(mcpTools[0].name).toBe('mcp__auth-server__authenticate')
-    expect(mcpTools[1].name).toBe('mcp__connected-server__do_thing')
-
-    // Reset mock for other tests
-    mockGetMcpToolsCommandsAndResources.mockReset()
-  })
-})
--- a/src/entrypoints/mcp.ts
+++ b/src/entrypoints/mcp.ts
@@ -7,7 +7,6 @@ process.env.CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS ??= 'true'

 import { Server } from '@modelcontextprotocol/sdk/server/index.js'
 import { StdioServerTransport } from '@modelcontextprotocol/sdk/server/stdio.js'
-import { ZodError } from 'zod'
 import {
  CallToolRequestSchema,
  type CallToolResult,
@@ -18,12 +17,9 @@ import {
 import { getDefaultAppState } from 'src/state/AppStateStore.js'
 import review from '../commands/review.js'
 import type { Command } from '../commands.js'
-import { getMcpToolsCommandsAndResources } from '../services/mcp/client.js'
-import type { MCPServerConnection } from '../services/mcp/types.js'
 import {
  findToolByName,
  getEmptyToolPermissionContext,
-  type Tool as InternalTool,
  type ToolUseContext,
 } from '../Tool.js'
 import { getTools } from '../tools.js'
@@ -43,32 +39,6 @@ type ToolOutput = Tool['outputSchema']

 const MCP_COMMANDS: Command[] = [review]

-export function getCombinedTools(
-  builtins: InternalTool[],
-  mcpTools: InternalTool[],
-): InternalTool[] {
-  const mcpToolNames = new Set(mcpTools.map(t => t.name))
-  const deduplicatedBuiltins = builtins.filter(t => !mcpToolNames.has(t.name))
-
-  return [...mcpTools, ...deduplicatedBuiltins]
-}
-
-export async function loadReexposedMcpTools(): Promise<{
-  mcpClients: MCPServerConnection[]
-  mcpTools: InternalTool[]
-}> {
-  const mcpClients: MCPServerConnection[] = []
-  const mcpTools: InternalTool[] = []
-
-  // Load configured MCP clients and their tools
-  await getMcpToolsCommandsAndResources(({ client, tools: clientTools }) => {
-    mcpClients.push(client)
-    mcpTools.push(...clientTools)
-  })
-
-  return { mcpClients, mcpTools }
-}
-
 export async function startMCPServer(
  cwd: string,
  debug: boolean,
@@ -93,13 +63,12 @@ export async function startMCPServer(
    },
  )

-  const { mcpClients, mcpTools } = await loadReexposedMcpTools()
-
  server.setRequestHandler(
    ListToolsRequestSchema,
    async (): Promise<ListToolsResult> => {
+      // TODO: Also re-expose any MCP tools
      const toolPermissionContext = getEmptyToolPermissionContext()
-      const tools = getCombinedTools(getTools(toolPermissionContext), mcpTools)
+      const tools = getTools(toolPermissionContext)
      return {
        tools: await Promise.all(
          tools.map(async tool => {
@@ -125,7 +94,7 @@ export async function startMCPServer(
                tools,
                agents: [],
              }),
-              inputSchema: (tool.inputJSONSchema ?? zodToJsonSchema(tool.inputSchema)) as ToolInput,
+              inputSchema: zodToJsonSchema(tool.inputSchema) as ToolInput,
              outputSchema,
            }
          }),
@@ -138,7 +107,8 @@ export async function startMCPServer(
    CallToolRequestSchema,
    async ({ params: { name, arguments: args } }): Promise<CallToolResult> => {
      const toolPermissionContext = getEmptyToolPermissionContext()
-      const tools = getCombinedTools(getTools(toolPermissionContext), mcpTools)
+      // TODO: Also re-expose any MCP tools
+      const tools = getTools(toolPermissionContext)
      const tool = findToolByName(tools, name)
      if (!tool) {
        throw new Error(`Tool ${name} not found`)
@@ -153,7 +123,7 @@ export async function startMCPServer(
          tools,
          mainLoopModel: getMainLoopModel(),
          thinkingConfig: { type: 'disabled' },
-          mcpClients,
+          mcpClients: [],
          mcpResources: {},
          isNonInteractiveSession: true,
          debug,
@@ -170,16 +140,13 @@ export async function startMCPServer(
        updateAttributionState: () => {},
      }

+      // TODO: validate input types with zod
      try {
        if (!tool.isEnabled()) {
          throw new Error(`Tool ${name} is not enabled`)
        }
-
-        // Validate input types with zod
-        const parsedArgs = tool.inputSchema.parse(args ?? {})
-
        const validationResult = await tool.validateInput?.(
-          (parsedArgs as never) ?? {},
+          (args as never) ?? {},
          toolUseContext,
        )
        if (validationResult && !validationResult.result) {
@@ -188,7 +155,7 @@ export async function startMCPServer(
          )
        }
        const finalResult = await tool.call(
-          (parsedArgs ?? {}) as never,
+          (args ?? {}) as never,
          toolUseContext,
          hasPermissionsToUseTool,
          createAssistantMessage({
@@ -196,50 +163,20 @@ export async function startMCPServer(
          }),
        )

-        let content: CallToolResult['content']
-        const data = finalResult.data as string | { type: string; text?: string; source?: { type: string; media_type: string; data: string } }[] | unknown
-
-        if (typeof data === 'string') {
-          content = [{ type: 'text', text: data }]
-        } else if (Array.isArray(data)) {
-          content = data.map((block: any) => {
-            if (block.type === 'text') {
-              return { type: 'text', text: block.text || '' }
-            } else if (block.type === 'image' && block.source) {
-              return {
-                type: 'image',
-                data: block.source.data,
-                mimeType: block.source.media_type,
-              }
-            } else {
-              // eslint-disable-next-line custom-rules/no-top-level-side-effects, no-console
-              console.warn(`Unmapped content block type from tool ${name}: ${block.type || 'unknown'}`)
-              return { type: 'text', text: jsonStringify(block) }
-            }
-          }) as CallToolResult['content']
-        } else {
-          content = [{ type: 'text', text: jsonStringify(data) }]
-        }
-
        return {
-          content,
-          isError: !!(finalResult as any).isError,
+          content: [
+            {
+              type: 'text' as const,
+              text:
+                typeof finalResult === 'string'
+                  ? finalResult
+                  : jsonStringify(finalResult.data),
+            },
+          ],
        }
      } catch (error) {
        logError(error)

-        if (error instanceof ZodError) {
-          return {
-            isError: true,
-            content: [
-              {
-                type: 'text',
-                text: `Tool ${name} input is invalid:\n${error.errors.map(e => `- ${e.path.join('.')}: ${e.message}`).join('\n')}`,
-              },
-            ],
-          }
-        }
-
        const parts =
          error instanceof Error ? getErrorParts(error) : [String(error)]
        const errorText = parts.filter(Boolean).join('\n').trim() || 'Error'
@@ -264,4 +201,3 @@ export async function startMCPServer(

  return await runServer()
 }
-
--- a/src/services/api/codexShim.ts
+++ b/src/services/api/codexShim.ts
@@ -1,5 +1,4 @@
 import { APIError } from '@anthropic-ai/sdk'
-import { fetchWithProxyRetry } from './fetchWithProxyRetry.js'
 import type {
  ResolvedCodexCredentials,
  ResolvedProviderRequest,
@@ -560,15 +559,12 @@ export async function performCodexRequest(options: {
  }
  headers.originator ??= 'openclaude'

-  const response = await fetchWithProxyRetry(
-    `${options.request.baseUrl}/responses`,
-    {
-      method: 'POST',
-      headers,
-      body: JSON.stringify(body),
-      signal: options.signal,
-    },
-  )
+  const response = await fetch(`${options.request.baseUrl}/responses`, {
+    method: 'POST',
+    headers,
+    body: JSON.stringify(body),
+    signal: options.signal,
+  })

  if (!response.ok) {
    const errorBody = await response.text().catch(() => 'unknown error')
--- a/src/services/api/errors.openaiCompatibility.test.ts
+++ b/src/services/api/errors.openaiCompatibility.test.ts
@@ -1,44 +0,0 @@
-import { APIError } from '@anthropic-ai/sdk'
-import { expect, test } from 'bun:test'
-
-import { getAssistantMessageFromError } from './errors.js'
-
-function getFirstText(message: ReturnType<typeof getAssistantMessageFromError>): string {
-  const first = message.message.content[0]
-  if (!first || typeof first !== 'object' || !('text' in first)) {
-    return ''
-  }
-  return typeof first.text === 'string' ? first.text : ''
-}
-
-test('maps endpoint_not_found category markers to actionable setup guidance', () => {
-  const error = APIError.generate(
-    404,
-    undefined,
-    'OpenAI API error 404: Not Found [openai_category=endpoint_not_found] Hint: Confirm OPENAI_BASE_URL includes /v1.',
-    new Headers(),
-  )
-
-  const message = getAssistantMessageFromError(error, 'qwen2.5-coder:7b')
-  const text = getFirstText(message)
-
-  expect(message.isApiErrorMessage).toBe(true)
-  expect(text).toContain('Provider endpoint was not found')
-  expect(text).toContain('OPENAI_BASE_URL')
-  expect(text).toContain('/v1')
-})
-
-test('maps tool_call_incompatible category markers to model/tool guidance', () => {
-  const error = APIError.generate(
-    400,
-    undefined,
-    'OpenAI API error 400: tool_calls are not supported [openai_category=tool_call_incompatible]',
-    new Headers(),
-  )
-
-  const message = getAssistantMessageFromError(error, 'qwen2.5-coder:7b')
-  const text = getFirstText(message)
-
-  expect(text).toContain('rejected tool-calling payloads')
-  expect(text).toContain('/model')
-})
--- a/src/services/api/errors.ts
+++ b/src/services/api/errors.ts
@@ -50,110 +50,9 @@ import {
 } from '../claudeAiLimits.js'
 import { shouldProcessRateLimits } from '../rateLimitMocking.js' // Used for /mock-limits command
 import { extractConnectionErrorDetails, formatAPIError } from './errorUtils.js'
-import {
-  extractOpenAICategoryMarker,
-  type OpenAICompatibilityFailureCategory,
-} from './openaiErrorClassification.js'

 export const API_ERROR_MESSAGE_PREFIX = 'API Error'

-function stripOpenAICompatibilityMetadata(message: string): string {
-  return message
-    .replace(/\s*\[openai_category=[a-z_]+\]\s*/g, ' ')
-    .replace(/\s{2,}/g, ' ')
-    .trim()
-}
-
-function mapOpenAICompatibilityFailureToAssistantMessage(options: {
-  category: OpenAICompatibilityFailureCategory
-  model: string
-  rawMessage: string
-}): AssistantMessage {
-  const switchCmd = getIsNonInteractiveSession() ? '--model' : '/model'
-  const compactHint = getIsNonInteractiveSession()
-    ? 'Reduce prompt size or start a new session.'
-    : 'Run /compact or start a new session with /new.'
-
-  switch (options.category) {
-    case 'localhost_resolution_failed':
-    case 'connection_refused':
-      return createAssistantAPIErrorMessage({
-        content:
-          'Could not connect to the local OpenAI-compatible provider. Ensure the local server is running, then use OPENAI_BASE_URL=http://127.0.0.1:11434/v1 for Ollama.',
-        error: 'unknown',
-      })
-
-    case 'endpoint_not_found':
-      return createAssistantAPIErrorMessage({
-        content:
-          'Provider endpoint was not found. Confirm OPENAI_BASE_URL targets an OpenAI-compatible /v1 endpoint (for Ollama: http://127.0.0.1:11434/v1).',
-        error: 'invalid_request',
-      })
-
-    case 'model_not_found':
-      return createAssistantAPIErrorMessage({
-        content: `The selected model (${options.model}) is not available on this provider. Run ${switchCmd} to choose another model, or verify installed local models (for Ollama: ollama list).`,
-        error: 'invalid_request',
-      })
-
-    case 'auth_invalid':
-      return createAssistantAPIErrorMessage({
-        content: `${API_ERROR_MESSAGE_PREFIX}: Authentication failed for your OpenAI-compatible provider. Verify OPENAI_API_KEY and endpoint-specific auth requirements.`,
-        error: 'authentication_failed',
-      })
-
-    case 'rate_limited':
-      return createAssistantAPIErrorMessage({
-        content: `${API_ERROR_MESSAGE_PREFIX}: Provider rate limit reached. Retry in a few seconds.`,
-        error: 'rate_limit',
-      })
-
-    case 'request_timeout':
-      return createAssistantAPIErrorMessage({
-        content: `${API_ERROR_MESSAGE_PREFIX}: Provider request timed out. Local models may be loading or overloaded; retry shortly or increase API_TIMEOUT_MS.`,
-        error: 'unknown',
-      })
-
-    case 'context_overflow':
-      return createAssistantAPIErrorMessage({
-        content: `The conversation exceeded the provider context limit. ${compactHint}`,
-        error: 'invalid_request',
-      })
-
-    case 'tool_call_incompatible':
-      return createAssistantAPIErrorMessage({
-        content: `The selected provider/model rejected tool-calling payloads. Try ${switchCmd} to pick a tool-capable model or continue without tools.`,
-        error: 'invalid_request',
-      })
-
-    case 'malformed_provider_response':
-      return createAssistantAPIErrorMessage({
-        content: `${API_ERROR_MESSAGE_PREFIX}: Provider returned a malformed response. Confirm endpoint compatibility and check local proxy/network middleware.`,
-        error: 'unknown',
-        errorDetails: stripOpenAICompatibilityMetadata(options.rawMessage),
-      })
-
-    case 'provider_unavailable':
-      return createAssistantAPIErrorMessage({
-        content: `${API_ERROR_MESSAGE_PREFIX}: Provider is temporarily unavailable. Retry in a moment.`,
-        error: 'unknown',
-      })
-
-    case 'network_error':
-    case 'unknown':
-      return createAssistantAPIErrorMessage({
-        content: `${API_ERROR_MESSAGE_PREFIX}: ${stripOpenAICompatibilityMetadata(options.rawMessage)}`,
-        error: 'unknown',
-      })
-
-    default:
-      return createAssistantAPIErrorMessage({
-        content: `${API_ERROR_MESSAGE_PREFIX}: ${stripOpenAICompatibilityMetadata(options.rawMessage)}`,
-        error: 'unknown',
-      })
-  }
-}
-
 export function startsWithApiErrorPrefix(text: string): boolean {
  return (
    text.startsWith(API_ERROR_MESSAGE_PREFIX) ||
@@ -558,19 +457,6 @@ export function getAssistantMessageFromError(
    })
  }

-  // OpenAI-compatible transport and HTTP failures include structured category
-  // markers from openaiShim.ts for actionable end-user remediation.
-  if (error instanceof APIError) {
-    const openaiCategory = extractOpenAICategoryMarker(error.message)
-    if (openaiCategory) {
-      return mapOpenAICompatibilityFailureToAssistantMessage({
-        category: openaiCategory,
-        model,
-        rawMessage: error.message,
-      })
-    }
-  }
-
  // Check for emergency capacity off switch for Opus PAYG users
  if (
    error instanceof Error &&
--- a/src/services/api/fetchWithProxyRetry.test.ts
+++ b/src/services/api/fetchWithProxyRetry.test.ts
@@ -1,86 +0,0 @@
-import { afterEach, beforeEach, expect, test } from 'bun:test'
-
-import { _resetKeepAliveForTesting } from '../../utils/proxy.js'
-import {
-  fetchWithProxyRetry,
-  isRetryableFetchError,
-} from './fetchWithProxyRetry.js'
-
-type FetchType = typeof globalThis.fetch
-
-const originalFetch = globalThis.fetch
-const originalEnv = {
-  HTTP_PROXY: process.env.HTTP_PROXY,
-  HTTPS_PROXY: process.env.HTTPS_PROXY,
-}
-
-function restoreEnv(key: 'HTTP_PROXY' | 'HTTPS_PROXY', value: string | undefined): void {
-  if (value === undefined) {
-    delete process.env[key]
-  } else {
-    process.env[key] = value
-  }
-}
-
-beforeEach(() => {
-  process.env.HTTP_PROXY = 'http://127.0.0.1:15236'
-  delete process.env.HTTPS_PROXY
-  _resetKeepAliveForTesting()
-})
-
-afterEach(() => {
-  globalThis.fetch = originalFetch
-  restoreEnv('HTTP_PROXY', originalEnv.HTTP_PROXY)
-  restoreEnv('HTTPS_PROXY', originalEnv.HTTPS_PROXY)
-  _resetKeepAliveForTesting()
-})
-
-test('isRetryableFetchError matches Bun socket-closed failures', () => {
-  expect(
-    isRetryableFetchError(
-      new Error(
-        'The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()',
-      ),
-    ),
-  ).toBe(true)
-})
-
-test('fetchWithProxyRetry retries once with keepalive disabled after socket closure', async () => {
-  const calls: Array<RequestInit | undefined> = []
-
-  globalThis.fetch = (async (_input, init) => {
-    calls.push(init)
-    if (calls.length === 1) {
-      throw new Error(
-        'The socket connection was closed unexpectedly. For more information, pass `verbose: true` in the second argument to fetch()',
-      )
-    }
-    return new Response('ok')
-  }) as FetchType
-
-  const response = await fetchWithProxyRetry('https://example.com/search', {
-    method: 'POST',
-  })
-
-  expect(await response.text()).toBe('ok')
-  expect(calls).toHaveLength(2)
-  expect((calls[0] as RequestInit & { proxy?: string }).proxy).toBe(
-    'http://127.0.0.1:15236',
-  )
-  expect((calls[0] as RequestInit).keepalive).toBeUndefined()
-  expect((calls[1] as RequestInit).keepalive).toBe(false)
-})
-
-test('fetchWithProxyRetry does not retry non-network errors', async () => {
-  let attempts = 0
-
-  globalThis.fetch = (async () => {
-    attempts += 1
-    throw new Error('400 bad request')
-  }) as FetchType
-
-  await expect(fetchWithProxyRetry('https://example.com')).rejects.toThrow(
-    '400 bad request',
-  )
-  expect(attempts).toBe(1)
-})
--- a/src/services/api/fetchWithProxyRetry.ts
+++ b/src/services/api/fetchWithProxyRetry.ts
@@ -1,44 +0,0 @@
-import { disableKeepAlive, getProxyFetchOptions } from '../../utils/proxy.js'
-
-const RETRYABLE_FETCH_ERROR_PATTERN =
-  /socket connection was closed unexpectedly|ECONNRESET|EPIPE|socket hang up|Connection reset by peer|fetch failed/i
-
-export function isRetryableFetchError(error: unknown): boolean {
-  if (!(error instanceof Error)) {
-    return false
-  }
-  if (error.name === 'AbortError') {
-    return false
-  }
-  return RETRYABLE_FETCH_ERROR_PATTERN.test(error.message)
-}
-
-export async function fetchWithProxyRetry(
-  input: string | URL | Request,
-  init?: RequestInit,
-  options?: { forAnthropicAPI?: boolean; maxAttempts?: number },
-): Promise<Response> {
-  const maxAttempts = Math.max(1, options?.maxAttempts ?? 2)
-  let lastError: unknown
-
-  for (let attempt = 1; attempt <= maxAttempts; attempt++) {
-    try {
-      return await fetch(input, {
-        ...init,
-        ...getProxyFetchOptions({
-          forAnthropicAPI: options?.forAnthropicAPI,
-        }),
-      })
-    } catch (error) {
-      lastError = error
-      if (attempt >= maxAttempts || !isRetryableFetchError(error)) {
-        throw error
-      }
-      disableKeepAlive()
-    }
-  }
-
-  throw lastError instanceof Error
-    ? lastError
-    : new Error('Fetch failed without an error object')
-}
--- a/src/services/api/openaiErrorClassification.test.ts
+++ b/src/services/api/openaiErrorClassification.test.ts
@@ -1,97 +0,0 @@
-import { expect, test } from 'bun:test'
-
-import {
-  buildOpenAICompatibilityErrorMessage,
-  classifyOpenAIHttpFailure,
-  classifyOpenAINetworkFailure,
-  extractOpenAICategoryMarker,
-  formatOpenAICategoryMarker,
-} from './openaiErrorClassification.js'
-
-test('classifies localhost ECONNREFUSED as connection_refused', () => {
-  const error = Object.assign(new TypeError('fetch failed'), {
-    code: 'ECONNREFUSED',
-  })
-
-  const failure = classifyOpenAINetworkFailure(error, {
-    url: 'http://localhost:11434/v1/chat/completions',
-  })
-
-  expect(failure.category).toBe('connection_refused')
-  expect(failure.retryable).toBe(true)
-  expect(failure.code).toBe('ECONNREFUSED')
-  expect(failure.hint).toContain('local server is running')
-})
-
-test('classifies localhost ENOTFOUND as localhost_resolution_failed', () => {
-  const error = Object.assign(new TypeError('getaddrinfo ENOTFOUND localhost'), {
-    code: 'ENOTFOUND',
-  })
-
-  const failure = classifyOpenAINetworkFailure(error, {
-    url: 'http://localhost:11434/v1/chat/completions',
-  })
-
-  expect(failure.category).toBe('localhost_resolution_failed')
-  expect(failure.retryable).toBe(true)
-  expect(failure.code).toBe('ENOTFOUND')
-  expect(failure.hint).toContain('127.0.0.1')
-})
-
-test('classifies model-not-found 404 responses', () => {
-  const failure = classifyOpenAIHttpFailure({
-    status: 404,
-    body: 'The model qwen2.5-coder:7b was not found',
-  })
-
-  expect(failure.category).toBe('model_not_found')
-  expect(failure.retryable).toBe(false)
-})
-
-test('classifies generic 404 responses as endpoint_not_found', () => {
-  const failure = classifyOpenAIHttpFailure({
-    status: 404,
-    body: 'Not Found',
-  })
-
-  expect(failure.category).toBe('endpoint_not_found')
-  expect(failure.hint).toContain('/v1')
-})
-
-test('classifies context-overflow responses', () => {
-  const failure = classifyOpenAIHttpFailure({
-    status: 500,
-    body: 'request too large: maximum context length exceeded',
-  })
-
-  expect(failure.category).toBe('context_overflow')
-  expect(failure.retryable).toBe(false)
-})
-
-test('classifies tool compatibility failures', () => {
-  const failure = classifyOpenAIHttpFailure({
-    status: 400,
-    body: 'tool_calls are not supported by this model',
-  })
-
-  expect(failure.category).toBe('tool_call_incompatible')
-})
-
-test('embeds and extracts category markers in formatted messages', () => {
-  const marker = formatOpenAICategoryMarker('endpoint_not_found')
-  expect(marker).toBe('[openai_category=endpoint_not_found]')
-
-  const formatted = buildOpenAICompatibilityErrorMessage('OpenAI API error 404: Not Found', {
-    category: 'endpoint_not_found',
-    hint: 'Confirm OPENAI_BASE_URL includes /v1.',
-  })
-
-  expect(formatted).toContain('[openai_category=endpoint_not_found]')
-  expect(formatted).toContain('Hint: Confirm OPENAI_BASE_URL includes /v1.')
-  expect(extractOpenAICategoryMarker(formatted)).toBe('endpoint_not_found')
-})
-
-test('ignores unknown category markers during extraction', () => {
-  const malformed = 'OpenAI API error 500 [openai_category=totally_fake_category]'
-  expect(extractOpenAICategoryMarker(malformed)).toBeUndefined()
-})
--- a/src/services/api/openaiErrorClassification.ts
+++ b/src/services/api/openaiErrorClassification.ts
@@ -1,355 +0,0 @@
-export type OpenAICompatibilityFailureCategory =
-  | 'connection_refused'
-  | 'localhost_resolution_failed'
-  | 'request_timeout'
-  | 'network_error'
-  | 'auth_invalid'
-  | 'rate_limited'
-  | 'model_not_found'
-  | 'endpoint_not_found'
-  | 'context_overflow'
-  | 'tool_call_incompatible'
-  | 'malformed_provider_response'
-  | 'provider_unavailable'
-  | 'unknown'
-
-export type OpenAICompatibilityFailure = {
-  source: 'network' | 'http'
-  category: OpenAICompatibilityFailureCategory
-  retryable: boolean
-  message: string
-  hint?: string
-  code?: string
-  status?: number
-}
-
-const OPENAI_CATEGORY_MARKER_PREFIX = '[openai_category='
-
-const LOCALHOST_HOSTNAMES = new Set(['localhost', '127.0.0.1', '::1'])
-
-const OPENAI_COMPATIBILITY_FAILURE_CATEGORIES: ReadonlySet<OpenAICompatibilityFailureCategory> =
-  new Set<OpenAICompatibilityFailureCategory>([
-    'connection_refused',
-    'localhost_resolution_failed',
-    'request_timeout',
-    'network_error',
-    'auth_invalid',
-    'rate_limited',
-    'model_not_found',
-    'endpoint_not_found',
-    'context_overflow',
-    'tool_call_incompatible',
-    'malformed_provider_response',
-    'provider_unavailable',
-    'unknown',
-  ])
-
-function isOpenAICompatibilityFailureCategory(
-  value: string,
-): value is OpenAICompatibilityFailureCategory {
-  return OPENAI_COMPATIBILITY_FAILURE_CATEGORIES.has(
-    value as OpenAICompatibilityFailureCategory,
-  )
-}
-
-function getErrorCode(error: unknown): string | undefined {
-  let current: unknown = error
-  const maxDepth = 5
-
-  for (let depth = 0; depth < maxDepth; depth++) {
-    if (
-      current &&
-      typeof current === 'object' &&
-      'code' in current &&
-      typeof (current as { code?: unknown }).code === 'string'
-    ) {
-      return (current as { code: string }).code
-    }
-
-    if (
-      current &&
-      typeof current === 'object' &&
-      'cause' in current &&
-      (current as { cause?: unknown }).cause !== current
-    ) {
-      current = (current as { cause?: unknown }).cause
-      continue
-    }
-
-    break
-  }
-
-  return undefined
-}
-
-function getHostname(url: string): string | null {
-  try {
-    return new URL(url).hostname.toLowerCase()
-  } catch {
-    return null
-  }
-}
-
-function isLocalhostLikeHostname(hostname: string | null): boolean {
-  if (!hostname) return false
-  if (LOCALHOST_HOSTNAMES.has(hostname)) return true
-  return /^127\./.test(hostname)
-}
-
-function isContextOverflowMessage(body: string): boolean {
-  const lower = body.toLowerCase()
-  return (
-    lower.includes('too many tokens') ||
-    lower.includes('request too large') ||
-    lower.includes('context length') ||
-    lower.includes('maximum context') ||
-    lower.includes('input length') ||
-    lower.includes('payload too large') ||
-    lower.includes('prompt is too long')
-  )
-}
-
-function isToolCompatibilityMessage(body: string): boolean {
-  const lower = body.toLowerCase()
-  return (
-    lower.includes('tool_calls') ||
-    lower.includes('tool_call') ||
-    lower.includes('tool_use') ||
-    lower.includes('tool_result') ||
-    lower.includes('function calling') ||
-    lower.includes('function call')
-  )
-}
-
-function isMalformedProviderResponse(body: string): boolean {
-  const lower = body.toLowerCase()
-  return (
-    lower.includes('<!doctype html') ||
-    lower.includes('<html') ||
-    lower.includes('invalid json') ||
-    lower.includes('malformed') ||
-    lower.includes('unexpected token') ||
-    lower.includes('cannot parse') ||
-    lower.includes('not valid json')
-  )
-}
-
-function isModelNotFoundMessage(body: string): boolean {
-  const lower = body.toLowerCase()
-  return (
-    lower.includes('model') &&
-    (
-      lower.includes('not found') ||
-      lower.includes('does not exist') ||
-      lower.includes('unknown model') ||
-      lower.includes('unavailable model')
-    )
-  )
-}
-
-export function formatOpenAICategoryMarker(
-  category: OpenAICompatibilityFailureCategory,
-): string {
-  return `${OPENAI_CATEGORY_MARKER_PREFIX}${category}]`
-}
-
-export function extractOpenAICategoryMarker(
-  message: string,
-): OpenAICompatibilityFailureCategory | undefined {
-  const match = message.match(/\[openai_category=([a-z_]+)]/)
-  const category = match?.[1]
-
-  if (!category || !isOpenAICompatibilityFailureCategory(category)) {
-    return undefined
-  }
-
-  return category
-}
-
-export function buildOpenAICompatibilityErrorMessage(
-  baseMessage: string,
-  failure: Pick<OpenAICompatibilityFailure, 'category' | 'hint'>,
-): string {
-  const marker = formatOpenAICategoryMarker(failure.category)
-  const hint = failure.hint ? ` Hint: ${failure.hint}` : ''
-  return `${baseMessage} ${marker}${hint}`
-}
-
-export function classifyOpenAINetworkFailure(
-  error: unknown,
-  options: { url: string },
-): OpenAICompatibilityFailure {
-  const message = error instanceof Error ? error.message : String(error)
-  const lowerMessage = message.toLowerCase()
-  const code = getErrorCode(error)
-  const hostname = getHostname(options.url)
-  const isLocalHost = isLocalhostLikeHostname(hostname)
-
-  if (
-    code === 'ETIMEDOUT' ||
-    code === 'UND_ERR_CONNECT_TIMEOUT' ||
-    lowerMessage.includes('timeout') ||
-    lowerMessage.includes('timed out') ||
-    lowerMessage.includes('aborterror')
-  ) {
-    return {
-      source: 'network',
-      category: 'request_timeout',
-      retryable: true,
-      message,
-      code,
-      hint: 'The provider took too long to respond. Check local model load time or increase API timeout.',
-    }
-  }
-
-  if (
-    isLocalHost &&
-    (
-      code === 'ENOTFOUND' ||
-      code === 'EAI_AGAIN' ||
-      lowerMessage.includes('getaddrinfo') ||
-      (code === undefined && lowerMessage.includes('fetch failed'))
-    )
-  ) {
-    return {
-      source: 'network',
-      category: 'localhost_resolution_failed',
-      retryable: true,
-      message,
-      code,
-      hint: 'Localhost failed for this request. Retry with 127.0.0.1 and confirm Ollama is serving on the configured port.',
-    }
-  }
-
-  if (code === 'ECONNREFUSED') {
-    return {
-      source: 'network',
-      category: 'connection_refused',
-      retryable: true,
-      message,
-      code,
-      hint: isLocalHost
-        ? 'Connection to the local provider was refused. Ensure the local server is running and listening on the configured port.'
-        : 'Connection was refused by the provider endpoint. Ensure the server is running and the port is correct.',
-    }
-  }
-
-  return {
-    source: 'network',
-    category: 'network_error',
-    retryable: true,
-    message,
-    code,
-    hint: 'Network transport failed before a provider response was received.',
-  }
-}
-
-export function classifyOpenAIHttpFailure(options: {
-  status: number
-  body: string
-}): OpenAICompatibilityFailure {
-  const body = options.body ?? ''
-
-  if (options.status === 401 || options.status === 403) {
-    return {
-      source: 'http',
-      category: 'auth_invalid',
-      retryable: false,
-      status: options.status,
-      message: body,
-      hint: 'Authentication failed. Verify API key, token source, and endpoint-specific auth headers.',
-    }
-  }
-
-  if (options.status === 429) {
-    return {
-      source: 'http',
-      category: 'rate_limited',
-      retryable: true,
-      status: options.status,
-      message: body,
-      hint: 'Provider rate-limited the request. Retry after backoff.',
-    }
-  }
-
-  if (options.status === 404 && isModelNotFoundMessage(body)) {
-    return {
-      source: 'http',
-      category: 'model_not_found',
-      retryable: false,
-      status: options.status,
-      message: body,
-      hint: 'The selected model is not installed or not available on this endpoint.',
-    }
-  }
-
-  if (options.status === 404) {
-    return {
-      source: 'http',
-      category: 'endpoint_not_found',
-      retryable: false,
-      status: options.status,
-      message: body,
-      hint: 'Endpoint was not found. Confirm OPENAI_BASE_URL includes /v1 for OpenAI-compatible local providers.',
-    }
-  }
-
-  if (
-    options.status === 413 ||
-    ((options.status === 400 || options.status >= 500) &&
-      isContextOverflowMessage(body))
-  ) {
-    return {
-      source: 'http',
-      category: 'context_overflow',
-      retryable: false,
-      status: options.status,
-      message: body,
-      hint: 'Prompt context exceeded model/server limits. Reduce context or increase provider context length.',
-    }
-  }
-
-  if (options.status === 400 && isToolCompatibilityMessage(body)) {
-    return {
-      source: 'http',
-      category: 'tool_call_incompatible',
-      retryable: false,
-      status: options.status,
-      message: body,
-      hint: 'Provider/model rejected tool-calling payload. Retry without tools or use a tool-capable model.',
-    }
-  }
-
-  if (
-    (options.status >= 200 && options.status < 300 && isMalformedProviderResponse(body)) ||
-    (options.status >= 400 && isMalformedProviderResponse(body))
-  ) {
-    return {
-      source: 'http',
-      category: 'malformed_provider_response',
-      retryable: false,
-      status: options.status,
-      message: body,
-      hint: 'Provider returned malformed or non-JSON response where JSON was expected.',
-    }
-  }
-
-  if (options.status >= 500) {
-    return {
-      source: 'http',
-      category: 'provider_unavailable',
-      retryable: true,
-      status: options.status,
-      message: body,
-      hint: 'Provider reported a server-side failure. Retry after a short delay.',
-    }
-  }
-
-  return {
-    source: 'http',
-    category: 'unknown',
-    retryable: false,
-    status: options.status,
-    message: body,
-  }
-}
--- a/src/services/api/openaiShim.diagnostics.test.ts
+++ b/src/services/api/openaiShim.diagnostics.test.ts
@@ -1,119 +0,0 @@
-import { afterEach, expect, mock, test } from 'bun:test'
-
-const originalFetch = globalThis.fetch
-const originalEnv = {
-  OPENAI_BASE_URL: process.env.OPENAI_BASE_URL,
-  OPENAI_API_KEY: process.env.OPENAI_API_KEY,
-  OPENAI_MODEL: process.env.OPENAI_MODEL,
-}
-
-function restoreEnv(key: string, value: string | undefined): void {
-  if (value === undefined) {
-    delete process.env[key]
-  } else {
-    process.env[key] = value
-  }
-}
-
-afterEach(() => {
-  globalThis.fetch = originalFetch
-  restoreEnv('OPENAI_BASE_URL', originalEnv.OPENAI_BASE_URL)
-  restoreEnv('OPENAI_API_KEY', originalEnv.OPENAI_API_KEY)
-  restoreEnv('OPENAI_MODEL', originalEnv.OPENAI_MODEL)
-  mock.restore()
-})
-
-test('logs classified transport diagnostics with category and code', async () => {
-  const debugSpy = mock(() => {})
-  mock.module('../../utils/debug.js', () => ({
-    logForDebugging: debugSpy,
-  }))
-
-  const nonce = `${Date.now()}-${Math.random()}`
-  const { createOpenAIShimClient } = await import(`./openaiShim.ts?ts=${nonce}`)
-
-  process.env.OPENAI_BASE_URL = 'http://localhost:11434/v1'
-  process.env.OPENAI_API_KEY = 'ollama'
-
-  const transportError = Object.assign(new TypeError('fetch failed'), {
-    code: 'ECONNREFUSED',
-  })
-
-  globalThis.fetch = mock(async () => {
-    throw transportError
-  }) as typeof globalThis.fetch
-
-  const client = createOpenAIShimClient({}) as {
-    beta: {
-      messages: {
-        create: (params: Record<string, unknown>) => Promise<unknown>
-      }
-    }
-  }
-
-  await expect(
-    client.beta.messages.create({
-      model: 'qwen2.5-coder:7b',
-      messages: [{ role: 'user', content: 'hello' }],
-      max_tokens: 64,
-      stream: false,
-    }),
-  ).rejects.toThrow('openai_category=connection_refused')
-
-  const transportLog = debugSpy.mock.calls.find(call =>
-    typeof call?.[0] === 'string' && call[0].includes('transport failure'),
-  )
-
-  expect(transportLog).toBeDefined()
-  expect(String(transportLog?.[0])).toContain('category=connection_refused')
-  expect(String(transportLog?.[0])).toContain('code=ECONNREFUSED')
-  expect(transportLog?.[1]).toEqual({ level: 'warn' })
-})
-
-test('redacts credentials in transport diagnostic URL logs', async () => {
-  const debugSpy = mock(() => {})
-  mock.module('../../utils/debug.js', () => ({
-    logForDebugging: debugSpy,
-  }))
-
-  const nonce = `${Date.now()}-${Math.random()}`
-  const { createOpenAIShimClient } = await import(`./openaiShim.ts?ts=${nonce}`)
-
-  process.env.OPENAI_BASE_URL = 'http://user:supersecret@localhost:11434/v1'
-  process.env.OPENAI_API_KEY = 'supersecret'
-
-  const transportError = Object.assign(new TypeError('fetch failed'), {
-    code: 'ECONNREFUSED',
-  })
-
-  globalThis.fetch = mock(async () => {
-    throw transportError
-  }) as typeof globalThis.fetch
-
-  const client = createOpenAIShimClient({}) as {
-    beta: {
-      messages: {
-        create: (params: Record<string, unknown>) => Promise<unknown>
-      }
-    }
-  }
-
-  await expect(
-    client.beta.messages.create({
-      model: 'qwen2.5-coder:7b',
-      messages: [{ role: 'user', content: 'hello' }],
-      max_tokens: 64,
-      stream: false,
-    }),
-  ).rejects.toThrow('openai_category=connection_refused')
-
-  const transportLog = debugSpy.mock.calls.find(call =>
-    typeof call?.[0] === 'string' && call[0].includes('transport failure'),
-  )
-
-  expect(transportLog).toBeDefined()
-  const logLine = String(transportLog?.[0])
-  expect(logLine).toContain('url=http://redacted:redacted@localhost:11434/v1/chat/completions')
-  expect(logLine).not.toContain('user:supersecret')
-  expect(logLine).not.toContain('supersecret@')
-})
--- a/src/services/api/openaiShim.test.ts
+++ b/src/services/api/openaiShim.test.ts
@@ -2775,172 +2775,3 @@ test('streaming: strips leaked reasoning preamble when split across multiple con

  expect(textDeltas).toEqual(['Hey! How can I help you today?'])
 })
-
-test('classifies localhost transport failures with actionable category marker', async () => {
-  process.env.OPENAI_BASE_URL = 'http://localhost:11434/v1'
-
-  const transportError = Object.assign(new TypeError('fetch failed'), {
-    code: 'ECONNREFUSED',
-  })
-
-  globalThis.fetch = (async () => {
-    throw transportError
-  }) as FetchType
-
-  const client = createOpenAIShimClient({}) as OpenAIShimClient
-
-  await expect(
-    client.beta.messages.create({
-      model: 'qwen2.5-coder:7b',
-      messages: [{ role: 'user', content: 'hello' }],
-      max_tokens: 64,
-      stream: false,
-    }),
-  ).rejects.toThrow('openai_category=connection_refused')
-
-  await expect(
-    client.beta.messages.create({
-      model: 'qwen2.5-coder:7b',
-      messages: [{ role: 'user', content: 'hello' }],
-      max_tokens: 64,
-      stream: false,
-    }),
-  ).rejects.toThrow('local server is running')
-})
-
-test('propagates AbortError without wrapping it as transport failure', async () => {
-  process.env.OPENAI_BASE_URL = 'http://localhost:11434/v1'
-
-  const abortError = new DOMException('The operation was aborted.', 'AbortError')
-  globalThis.fetch = (async () => {
-    throw abortError
-  }) as FetchType
-
-  const controller = new AbortController()
-  controller.abort()
-
-  const client = createOpenAIShimClient({}) as OpenAIShimClient
-
-  await expect(
-    client.beta.messages.create(
-      {
-        model: 'qwen2.5-coder:7b',
-        messages: [{ role: 'user', content: 'hello' }],
-        max_tokens: 64,
-        stream: false,
-      },
-      { signal: controller.signal },
-    ),
-  ).rejects.toBe(abortError)
-})
-
-test('classifies chat-completions endpoint 404 failures with endpoint_not_found marker', async () => {
-  process.env.OPENAI_BASE_URL = 'http://localhost:11434'
-
-  globalThis.fetch = (async () =>
-    new Response('Not Found', {
-      status: 404,
-      headers: {
-        'Content-Type': 'text/plain',
-      },
-    })) as FetchType
-
-  const client = createOpenAIShimClient({}) as OpenAIShimClient
-
-  await expect(
-    client.beta.messages.create({
-      model: 'qwen2.5-coder:7b',
-      messages: [{ role: 'user', content: 'hello' }],
-      max_tokens: 64,
-      stream: false,
-    }),
-  ).rejects.toThrow('openai_category=endpoint_not_found')
-})
-
-test('preserves valid tool_result and drops orphan tool_result', async () => {
-  let requestBody: Record<string, unknown> | undefined
-
-  globalThis.fetch = (async (_input, init) => {
-    requestBody = JSON.parse(String(init?.body))
-
-    return new Response(
-      JSON.stringify({
-        id: 'chatcmpl-1',
-        model: 'mistral-large-latest',
-        choices: [
-          {
-            message: {
-              role: 'assistant',
-              content: 'done',
-            },
-            finish_reason: 'stop',
-          },
-        ],
-        usage: {
-          prompt_tokens: 12,
-          completion_tokens: 4,
-          total_tokens: 16,
-        },
-      }),
-      {
-        headers: {
-          'Content-Type': 'application/json',
-        },
-      },
-    )
-  }) as FetchType
-
-  const client = createOpenAIShimClient({}) as OpenAIShimClient
-
-  await client.beta.messages.create({
-    model: 'mistral-large-latest',
-    system: 'test system',
-    messages: [
-      { role: 'user', content: 'Search and then I will interrupt' },
-      {
-        role: 'assistant',
-        content: [
-          {
-            type: 'tool_use',
-            id: 'valid_call_1',
-            name: 'Search',
-            input: { query: 'openclaude' },
-          },
-        ],
-      },
-      {
-        role: 'user',
-        content: [
-          {
-            type: 'tool_result',
-            tool_use_id: 'valid_call_1',
-            content: 'Found it!',
-          },
-          {
-            type: 'tool_result',
-            tool_use_id: 'orphan_call_2',
-            content: 'Interrupted result',
-          },
-          {
-            role: 'user',
-            content: 'What happened?',
-          }
-        ],
-      },
-    ],
-    max_tokens: 64,
-    stream: false,
-  })
-
-  const messages = requestBody?.messages as Array<Record<string, unknown>>
-  
-  // Should have: system, user, assistant (tool_use), tool (valid_call_1), user
-  // Should NOT have: tool (orphan_call_2)
-  
-  const toolMessages = messages.filter(m => m.role === 'tool')
-  expect(toolMessages.length).toBe(1)
-  expect(toolMessages[0].tool_call_id).toBe('valid_call_1')
-  
-  const orphanMessage = toolMessages.find(m => m.tool_call_id === 'orphan_call_2')
-  expect(orphanMessage).toBeUndefined()
-})
--- a/src/services/api/openaiShim.ts
+++ b/src/services/api/openaiShim.ts
@@ -47,18 +47,12 @@ import {
  type AnthropicUsage,
  type ShimCreateParams,
 } from './codexShim.js'
-import { fetchWithProxyRetry } from './fetchWithProxyRetry.js'
 import {
  isLocalProviderUrl,
  resolveRuntimeCodexCredentials,
  resolveProviderRequest,
  getGithubEndpointType,
 } from './providerConfig.js'
-import {
-  buildOpenAICompatibilityErrorMessage,
-  classifyOpenAIHttpFailure,
-  classifyOpenAINetworkFailure,
-} from './openaiErrorClassification.js'
 import { sanitizeSchemaForOpenAICompat } from '../../utils/schemaSanitizer.js'
 import { redactSecretValueForDisplay } from '../../utils/providerProfile.js'
 import {
@@ -88,19 +82,6 @@ const COPILOT_HEADERS: Record<string, string> = {
  'Copilot-Integration-Id': 'vscode-chat',
 }

-const SENSITIVE_URL_QUERY_PARAM_NAMES = [
-  'api_key',
-  'key',
-  'token',
-  'access_token',
-  'refresh_token',
-  'signature',
-  'sig',
-  'secret',
-  'password',
-  'authorization',
-]
-
 function isGithubModelsMode(): boolean {
  return isEnvTruthy(process.env.CLAUDE_CODE_USE_GITHUB)
 }
@@ -150,34 +131,6 @@ function formatRetryAfterHint(response: Response): string {
  return ra ? ` (Retry-After: ${ra})` : ''
 }

-function shouldRedactUrlQueryParam(name: string): boolean {
-  const lower = name.toLowerCase()
-  return SENSITIVE_URL_QUERY_PARAM_NAMES.some(token => lower.includes(token))
-}
-
-function redactUrlForDiagnostics(url: string): string {
-  try {
-    const parsed = new URL(url)
-    if (parsed.username) {
-      parsed.username = 'redacted'
-    }
-    if (parsed.password) {
-      parsed.password = 'redacted'
-    }
-
-    for (const key of parsed.searchParams.keys()) {
-      if (shouldRedactUrlQueryParam(key)) {
-        parsed.searchParams.set(key, 'redacted')
-      }
-    }
-
-    const serialized = parsed.toString()
-    return redactSecretValueForDisplay(serialized, process.env as SecretValueSource) ?? serialized
-  } catch {
-    return redactSecretValueForDisplay(url, process.env as SecretValueSource) ?? url
-  }
-}
-
 function sleepMs(ms: number): Promise<void> {
  return new Promise(resolve => setTimeout(resolve, ms))
 }
@@ -349,7 +302,6 @@ function convertMessages(
  system: unknown,
 ): OpenAIMessage[] {
  const result: OpenAIMessage[] = []
-  const knownToolCallIds = new Set<string>()

  // System message first
  const sysText = convertSystemPrompt(system)
@@ -369,21 +321,13 @@ function convertMessages(
        const toolResults = content.filter((b: { type?: string }) => b.type === 'tool_result')
        const otherContent = content.filter((b: { type?: string }) => b.type !== 'tool_result')

-        // Emit tool results as tool messages, but ONLY if we have a matching tool_use ID.
-        // Mistral/OpenAI strictly require tool messages to follow an assistant message with tool_calls.
-        // If the user interrupted (ESC) and a synthetic tool_result was generated without a recorded tool_use,
-        // emitting it here would cause a "role must alternate" or "unexpected role" error.
+        // Emit tool results as tool messages
        for (const tr of toolResults) {
-          const id = tr.tool_use_id ?? 'unknown'
-          if (knownToolCallIds.has(id)) {
-            result.push({
-              role: 'tool',
-              tool_call_id: id,
-              content: convertToolResultContent(tr.content, tr.is_error),
-            })
-          } else {
-            logForDebugging(`Dropping orphan tool_result for ID: ${id} to prevent API error`)
-          }
+          result.push({
+            role: 'tool',
+            tool_call_id: tr.tool_use_id ?? 'unknown',
+            content: convertToolResultContent(tr.content, tr.is_error),
+          })
        }

        // Emit remaining user content
@@ -424,11 +368,9 @@ function convertMessages(
              input?: unknown
              extra_content?: Record<string, unknown>
              signature?: string
-            }) => {
-              const id = tu.id ?? `call_${crypto.randomUUID().replace(/-/g, '')}`
-              knownToolCallIds.add(id)
+            }, index) => {
              const toolCall: NonNullable<OpenAIMessage['tool_calls']>[number] = {
-                id,
+                id: tu.id ?? `call_${crypto.randomUUID().replace(/-/g, '')}`,
                type: 'function' as const,
                function: {
                  name: tu.name ?? 'unknown',
@@ -453,6 +395,7 @@ function convertMessages(

                // Merge into existing google-specific metadata if present
                const existingGoogle = (toolCall.extra_content?.google as Record<string, unknown>) ?? {}
+
                toolCall.extra_content = {
                  ...toolCall.extra_content,
                  google: {
@@ -607,10 +550,7 @@ function convertTools(
        function: {
          name: t.name,
          description: t.description ?? '',
-          parameters: normalizeSchemaForOpenAI(
-            schema,
-            !isGemini && !isEnvTruthy(process.env.OPENCLAUDE_DISABLE_STRICT_TOOLS),
-          ),
+          parameters: normalizeSchemaForOpenAI(schema, !isGemini),
        },
      }
    })
@@ -1420,12 +1360,8 @@ class OpenAIShimMessages {
      ...filterAnthropicHeaders(options?.headers),
    }

-    const isGemini = isGeminiMode()
-    const isMiniMax = !!process.env.MINIMAX_API_KEY
-    const apiKey =
-      this.providerOverride?.apiKey ??
-      process.env.OPENAI_API_KEY ??
-      (isMiniMax ? process.env.MINIMAX_API_KEY : '')
+    const isGemini = isEnvTruthy(process.env.CLAUDE_CODE_USE_GEMINI)
+    const apiKey = this.providerOverride?.apiKey ?? process.env.OPENAI_API_KEY ?? ''
    // Detect Azure endpoints by hostname (not raw URL) to prevent bypass via
    // path segments like https://evil.com/cognitiveservices.azure.com/
    let isAzure = false
@@ -1489,97 +1425,12 @@ class OpenAIShimMessages {
    }

    const maxAttempts = isGithub ? GITHUB_429_MAX_RETRIES : 1
-
-    const throwClassifiedTransportError = (
-      error: unknown,
-      requestUrl: string,
-    ): never => {
-      if (options?.signal?.aborted) {
-        throw error
-      }
-
-      const failure = classifyOpenAINetworkFailure(error, {
-        url: requestUrl,
-      })
-      const redactedUrl = redactUrlForDiagnostics(requestUrl)
-      const safeMessage =
-        redactSecretValueForDisplay(
-          failure.message,
-          process.env as SecretValueSource,
-        ) || 'Request failed'
-
-      logForDebugging(
-        `[OpenAIShim] transport failure category=${failure.category} retryable=${failure.retryable} code=${failure.code ?? 'unknown'} method=POST url=${redactedUrl} model=${request.resolvedModel} message=${safeMessage}`,
-        { level: 'warn' },
-      )
-
-      throw APIError.generate(
-        503,
-        undefined,
-        buildOpenAICompatibilityErrorMessage(
-          `OpenAI API transport error: ${safeMessage}${failure.code ? ` (code=${failure.code})` : ''}`,
-          failure,
-        ),
-        new Headers(),
-      )
-    }
-
-    const throwClassifiedHttpError = (
-      status: number,
-      errorBody: string,
-      parsedBody: object | undefined,
-      responseHeaders: Headers,
-      requestUrl: string,
-      rateHint = '',
-    ): never => {
-      const failure = classifyOpenAIHttpFailure({
-        status,
-        body: errorBody,
-      })
-      const redactedUrl = redactUrlForDiagnostics(requestUrl)
-
-      logForDebugging(
-        `[OpenAIShim] request failed category=${failure.category} retryable=${failure.retryable} status=${status} method=POST url=${redactedUrl} model=${request.resolvedModel}`,
-        { level: 'warn' },
-      )
-
-      throw APIError.generate(
-        status,
-        parsedBody,
-        buildOpenAICompatibilityErrorMessage(
-          `OpenAI API error ${status}: ${errorBody}${rateHint}`,
-          failure,
-        ),
-        responseHeaders,
-      )
-    }
-
    let response: Response | undefined
    for (let attempt = 0; attempt < maxAttempts; attempt++) {
-      try {
-        response = await fetchWithProxyRetry(chatCompletionsUrl, fetchInit)
-      } catch (error) {
-        const isAbortError =
-          fetchInit.signal?.aborted === true ||
-          (typeof DOMException !== 'undefined' &&
-            error instanceof DOMException &&
-            error.name === 'AbortError') ||
-          (typeof error === 'object' &&
-            error !== null &&
-            'name' in error &&
-            error.name === 'AbortError')
-
-        if (isAbortError) {
-          throw error
-        }
-
-        throwClassifiedTransportError(error, chatCompletionsUrl)
-      }
-
+      response = await fetch(chatCompletionsUrl, fetchInit)
      if (response.ok) {
        return response
      }
-
      if (
        isGithub &&
        response.status === 429 &&
@@ -1649,43 +1500,34 @@ class OpenAIShimMessages {
            }
          }

-          let responsesResponse: Response
-          try {
-            responsesResponse = await fetchWithProxyRetry(responsesUrl, {
-              method: 'POST',
-              headers,
-              body: JSON.stringify(responsesBody),
-              signal: options?.signal,
-            })
-          } catch (error) {
-            throwClassifiedTransportError(error, responsesUrl)
-          }
-
+          const responsesResponse = await fetch(responsesUrl, {
+            method: 'POST',
+            headers,
+            body: JSON.stringify(responsesBody),
+            signal: options?.signal,
+          })
          if (responsesResponse.ok) {
            return responsesResponse
          }
          const responsesErrorBody = await responsesResponse.text().catch(() => 'unknown error')
          let responsesErrorResponse: object | undefined
          try { responsesErrorResponse = JSON.parse(responsesErrorBody) } catch { /* raw text */ }
-          throwClassifiedHttpError(
+          throw APIError.generate(
            responsesResponse.status,
-            responsesErrorBody,
            responsesErrorResponse,
+            `OpenAI API error ${responsesResponse.status}: ${responsesErrorBody}`,
            responsesResponse.headers,
-            responsesUrl,
          )
        }
      }

      let errorResponse: object | undefined
      try { errorResponse = JSON.parse(errorBody) } catch { /* raw text */ }
-      throwClassifiedHttpError(
+      throw APIError.generate(
        response.status,
-        errorBody,
        errorResponse,
+        `OpenAI API error ${response.status}: ${errorBody}${rateHint}`,
        response.headers as unknown as Headers,
-        chatCompletionsUrl,
-        rateHint,
      )
    }

--- a/src/services/api/providerConfig.envDiagnostics.test.ts
+++ b/src/services/api/providerConfig.envDiagnostics.test.ts
@@ -1,107 +0,0 @@
-import { afterEach, expect, mock, test } from 'bun:test'
-
-const originalEnv = {
-  CLAUDE_CODE_USE_OPENAI: process.env.CLAUDE_CODE_USE_OPENAI,
-  CLAUDE_CODE_USE_MISTRAL: process.env.CLAUDE_CODE_USE_MISTRAL,
-  OPENAI_BASE_URL: process.env.OPENAI_BASE_URL,
-  OPENAI_MODEL: process.env.OPENAI_MODEL,
-  OPENAI_API_BASE: process.env.OPENAI_API_BASE,
-  MISTRAL_BASE_URL: process.env.MISTRAL_BASE_URL,
-  MISTRAL_MODEL: process.env.MISTRAL_MODEL,
-}
-
-function restoreEnv(key: string, value: string | undefined): void {
-  if (value === undefined) {
-    delete process.env[key]
-  } else {
-    process.env[key] = value
-  }
-}
-
-afterEach(() => {
-  restoreEnv('CLAUDE_CODE_USE_OPENAI', originalEnv.CLAUDE_CODE_USE_OPENAI)
-  restoreEnv('CLAUDE_CODE_USE_MISTRAL', originalEnv.CLAUDE_CODE_USE_MISTRAL)
-  restoreEnv('OPENAI_BASE_URL', originalEnv.OPENAI_BASE_URL)
-  restoreEnv('OPENAI_MODEL', originalEnv.OPENAI_MODEL)
-  restoreEnv('OPENAI_API_BASE', originalEnv.OPENAI_API_BASE)
-  restoreEnv('MISTRAL_BASE_URL', originalEnv.MISTRAL_BASE_URL)
-  restoreEnv('MISTRAL_MODEL', originalEnv.MISTRAL_MODEL)
-  mock.restore()
-})
-
-test('logs a warning when OPENAI_BASE_URL is literal undefined', async () => {
-  const debugSpy = mock(() => {})
-  mock.module('../../utils/debug.js', () => ({
-    logForDebugging: debugSpy,
-  }))
-
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  process.env.OPENAI_BASE_URL = 'undefined'
-  process.env.OPENAI_MODEL = 'gpt-4o'
-  delete process.env.OPENAI_API_BASE
-
-  const nonce = `${Date.now()}-${Math.random()}`
-  const { resolveProviderRequest } = await import(`./providerConfig.ts?ts=${nonce}`)
-
-  const resolved = resolveProviderRequest()
-
-  expect(resolved.baseUrl).toBe('https://api.openai.com/v1')
-
-  const warningCall = debugSpy.mock.calls.find(call =>
-    typeof call?.[0] === 'string' &&
-    call[0].includes('OPENAI_BASE_URL') &&
-    call[0].includes('"undefined"'),
-  )
-
-  expect(warningCall).toBeDefined()
-  expect(warningCall?.[1]).toEqual({ level: 'warn' })
-})
-
-test('does not warn for OPENAI_API_BASE when OPENAI_BASE_URL is active', async () => {
-  const debugSpy = mock(() => {})
-  mock.module('../../utils/debug.js', () => ({
-    logForDebugging: debugSpy,
-  }))
-
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_USE_MISTRAL
-  process.env.OPENAI_BASE_URL = 'http://127.0.0.1:11434/v1'
-  process.env.OPENAI_MODEL = 'qwen2.5-coder:7b'
-  process.env.OPENAI_API_BASE = 'undefined'
-
-  const nonce = `${Date.now()}-${Math.random()}`
-  const { resolveProviderRequest } = await import(`./providerConfig.ts?ts=${nonce}`)
-
-  const resolved = resolveProviderRequest()
-
-  expect(resolved.baseUrl).toBe('http://127.0.0.1:11434/v1')
-
-  const aliasWarning = debugSpy.mock.calls.find(call =>
-    typeof call?.[0] === 'string' &&
-    call[0].includes('OPENAI_API_BASE') &&
-    call[0].includes('"undefined"'),
-  )
-
-  expect(aliasWarning).toBeUndefined()
-})
-
-test('uses OPENAI_API_BASE as fallback in mistral mode when MISTRAL_BASE_URL is unset', async () => {
-  const debugSpy = mock(() => {})
-  mock.module('../../utils/debug.js', () => ({
-    logForDebugging: debugSpy,
-  }))
-
-  delete process.env.CLAUDE_CODE_USE_OPENAI
-  process.env.CLAUDE_CODE_USE_MISTRAL = '1'
-  delete process.env.MISTRAL_BASE_URL
-  process.env.MISTRAL_MODEL = 'mistral-medium-latest'
-  process.env.OPENAI_API_BASE = 'http://127.0.0.1:11434/v1'
-
-  const nonce = `${Date.now()}-${Math.random()}`
-  const { resolveProviderRequest } = await import(`./providerConfig.ts?ts=${nonce}`)
-
-  const resolved = resolveProviderRequest()
-
-  expect(resolved.baseUrl).toBe('http://127.0.0.1:11434/v1')
-  expect(debugSpy.mock.calls).toHaveLength(0)
-})
--- a/src/services/api/providerConfig.ts
+++ b/src/services/api/providerConfig.ts
@@ -8,20 +8,17 @@ import {
  readCodexCredentials,
  type CodexCredentialBlob,
 } from '../../utils/codexCredentials.js'
-import { logForDebugging } from '../../utils/debug.js'
 import { isEnvTruthy } from '../../utils/envUtils.js'
 import {
  asTrimmedString,
  parseChatgptAccountId,
 } from './codexOAuthShared.js'
-import { DEFAULT_GEMINI_BASE_URL } from 'src/utils/providerProfile.js'

 export const DEFAULT_OPENAI_BASE_URL = 'https://api.openai.com/v1'
 export const DEFAULT_CODEX_BASE_URL = 'https://chatgpt.com/backend-api/codex'
 export const DEFAULT_MISTRAL_BASE_URL = 'https://api.mistral.ai/v1'
 /** Default GitHub Copilot API model when user selects copilot / github:copilot */
 export const DEFAULT_GITHUB_MODELS_API_MODEL = 'gpt-4o'
-const warnedUndefinedEnvNames = new Set<string>()

 const CODEX_ALIAS_MODELS: Record<
  string,
@@ -132,33 +129,7 @@ function isPrivateIpv6Address(hostname: string): boolean {
 function asEnvUrl(value: string | undefined): string | undefined {
  if (!value) return undefined
  const trimmed = value.trim()
-  if (!trimmed) return undefined
-  if (trimmed === 'undefined') {
-    return undefined
-  }
-  return trimmed
-}
-
-function asNamedEnvUrl(
-  value: string | undefined,
-  envName: string,
-): string | undefined {
-  if (!value) return undefined
-
-  const trimmed = value.trim()
-  if (!trimmed) return undefined
-
-  if (trimmed === 'undefined') {
-    if (!warnedUndefinedEnvNames.has(envName)) {
-      warnedUndefinedEnvNames.add(envName)
-      logForDebugging(
-        `[provider-config] Environment variable ${envName} is the literal string "undefined"; ignoring it.`,
-        { level: 'warn' },
-      )
-    }
-    return undefined
-  }
-
+  if (!trimmed || trimmed === 'undefined') return undefined
  return trimmed
 }

@@ -382,52 +353,23 @@ export function resolveProviderRequest(options?: {
 }): ResolvedProviderRequest {
  const isGithubMode = isEnvTruthy(process.env.CLAUDE_CODE_USE_GITHUB)
  const isMistralMode = isEnvTruthy(process.env.CLAUDE_CODE_USE_MISTRAL)
-  const isGeminiMode = isEnvTruthy(process.env.CLAUDE_CODE_USE_GEMINI)
  const requestedModel =
    options?.model?.trim() ||
    (isMistralMode
      ? process.env.MISTRAL_MODEL?.trim()
      : process.env.OPENAI_MODEL?.trim()) ||
-    (isGeminiMode
-      ? process.env.GEMINI_MODEL?.trim()
-      : process.env.OPENAI_MODEL?.trim()) ||
    options?.fallbackModel?.trim() ||
    (isGithubMode ? 'github:copilot' : 'gpt-4o')
  const descriptor = parseModelDescriptor(requestedModel)
  const explicitBaseUrl = asEnvUrl(options?.baseUrl)
-
-  const normalizedMistralEnvBaseUrl = asNamedEnvUrl(
-    process.env.MISTRAL_BASE_URL,
-    'MISTRAL_BASE_URL',
-  )
-
-  const normalizedGeminiEnvBaseUrl = asNamedEnvUrl(
-    process.env.GEMINI_BASE_URL,
-    'GEMINI_BASE_URL',
-  )
-
-  const primaryEnvBaseUrl = isMistralMode
-    ? normalizedMistralEnvBaseUrl
-    : isGeminiMode
-    ? normalizedGeminiEnvBaseUrl
-    : asNamedEnvUrl(process.env.OPENAI_BASE_URL, 'OPENAI_BASE_URL')
-
-  const fallbackEnvBaseUrl = isMistralMode
-    ? (primaryEnvBaseUrl === undefined
-      ? asNamedEnvUrl(process.env.OPENAI_API_BASE, 'OPENAI_API_BASE') ?? DEFAULT_MISTRAL_BASE_URL
-      : undefined)
-    : isGeminiMode
-    ? (primaryEnvBaseUrl === undefined
-      ? asNamedEnvUrl(process.env.OPENAI_API_BASE, 'OPENAI_API_BASE') ?? DEFAULT_GEMINI_BASE_URL
-      : undefined)
-    : (primaryEnvBaseUrl === undefined
-      ? asNamedEnvUrl(process.env.OPENAI_API_BASE, 'OPENAI_API_BASE')
-      : undefined)
-
  const envBaseUrlRaw =
    explicitBaseUrl ??
-    primaryEnvBaseUrl ??
-    fallbackEnvBaseUrl
+    asEnvUrl(
+      isMistralMode
+        ? (process.env.MISTRAL_BASE_URL ?? DEFAULT_MISTRAL_BASE_URL)
+        : process.env.OPENAI_BASE_URL
+    ) ??
+    asEnvUrl(process.env.OPENAI_API_BASE)

  const isCodexModelForGithub = isGithubMode && isCodexAlias(requestedModel)
  const envBaseUrl =
--- a/src/services/compact/autoCompact.ts
+++ b/src/services/compact/autoCompact.ts
@@ -110,14 +110,9 @@ export function calculateTokenWarningState(
    ? autoCompactThreshold
    : getEffectiveContextWindowSize(model)

-  // Use the raw context window (without output reservation) for the percentage
-  // display, so users see remaining context relative to the model's full capacity.
-  // The threshold (which subtracts buffer) should only affect when we warn/compact,
-  // not what percentage we display.
-  const rawContextWindow = getContextWindowForModel(model, getSdkBetas())
  const percentLeft = Math.max(
    0,
-    Math.round(((rawContextWindow - tokenUsage) / rawContextWindow) * 100),
+    Math.round(((threshold - tokenUsage) / threshold) * 100),
  )

  const warningThreshold = threshold - WARNING_THRESHOLD_BUFFER_TOKENS
--- a/src/services/diagnosticTracking.test.ts
+++ b/src/services/diagnosticTracking.test.ts
@@ -1,152 +0,0 @@
-import { describe, test, expect, beforeEach, afterEach } from 'bun:test'
-import { DiagnosticTrackingService } from './diagnosticTracking.js'
-import type { MCPServerConnection } from './mcp/types.js'
-
-// Mock the IDE client utility
-const mockGetConnectedIdeClient = (clients: MCPServerConnection[]) => 
-  clients.find(client => client.type === 'connected')
-
-describe('DiagnosticTrackingService', () => {
-  let service: DiagnosticTrackingService
-  let mockClients: MCPServerConnection[]
-  let mockIdeClient: MCPServerConnection
-
-  beforeEach(() => {
-    // Get fresh instance for each test
-    service = DiagnosticTrackingService.getInstance()
-    
-    // Setup mock clients
-    mockIdeClient = {
-      type: 'connected',
-      name: 'test-ide',
-      capabilities: {},
-      config: {},
-      cleanup: async () => {},
-      client: {
-        request: async () => ({}),
-        setNotificationHandler: () => {},
-        close: async () => {},
-      },
-    } as unknown as MCPServerConnection
-
-    mockClients = [
-      { type: 'disconnected', name: 'test-disconnected', config: {} } as unknown as MCPServerConnection,
-      mockIdeClient,
-    ]
-  })
-
-  afterEach(async () => {
-    await service.shutdown()
-  })
-
-  describe('handleQueryStart', () => {
-    test('should store MCP clients and initialize service', async () => {
-      await service.handleQueryStart(mockClients)
-
-      // Service should be initialized
-      expect(service).toBeDefined()
-
-      // Should be able to get IDE client from stored clients
-      // We can't directly test private methods, but we can test the behavior
-      const result = await service.getNewDiagnosticsCompat()
-      expect(result).toEqual([]) // Should return empty when no diagnostics
-    })
-
-    test('should reset service if already initialized', async () => {
-      // Initialize first
-      await service.handleQueryStart(mockClients)
-      
-      // Call again - should reset without error
-      await service.handleQueryStart(mockClients)
-      
-      // Should still work
-      const result = await service.getNewDiagnosticsCompat()
-      expect(result).toEqual([])
-    })
-  })
-
-  describe('backward-compatible methods', () => {
-    beforeEach(async () => {
-      await service.handleQueryStart(mockClients)
-    })
-
-    test('beforeFileEditedCompat should work without explicit client', async () => {
-      // Should not throw error and should return undefined when no IDE client
-      const result = await service.beforeFileEditedCompat('/test/file.ts')
-      expect(result).toBeUndefined()
-    })
-
-    test('getNewDiagnosticsCompat should work without explicit client', async () => {
-      const result = await service.getNewDiagnosticsCompat()
-      expect(Array.isArray(result)).toBe(true)
-    })
-
-    test('ensureFileOpenedCompat should work without explicit client', async () => {
-      const result = await service.ensureFileOpenedCompat('/test/file.ts')
-      expect(result).toBeUndefined()
-    })
-  })
-
-  describe('new explicit client methods', () => {
-    test('beforeFileEdited should require client parameter', async () => {
-      // Should not work without client
-      const result = await service.beforeFileEdited('/test/file.ts', undefined as any)
-      expect(result).toBeUndefined()
-    })
-
-    test('getNewDiagnostics should require client parameter', async () => {
-      // Should not work without client
-      const result = await service.getNewDiagnostics(undefined as any)
-      expect(result).toEqual([])
-    })
-
-    test('ensureFileOpened should require client parameter', async () => {
-      // Should not work without client
-      const result = await service.ensureFileOpened('/test/file.ts', undefined as any)
-      expect(result).toBeUndefined()
-    })
-  })
-
-  describe('shutdown', () => {
-    test('should clear stored clients on shutdown', async () => {
-      await service.handleQueryStart(mockClients)
-      
-      // Verify service is working
-      const beforeResult = await service.getNewDiagnosticsCompat()
-      expect(Array.isArray(beforeResult)).toBe(true)
-      
-      // Shutdown
-      await service.shutdown()
-      
-      // After shutdown, compat methods should return empty results
-      const afterResult = await service.getNewDiagnosticsCompat()
-      expect(afterResult).toEqual([])
-    })
-  })
-
-  describe('integration with existing functionality', () => {
-    test('should maintain existing diagnostic tracking behavior', async () => {
-      await service.handleQueryStart(mockClients)
-      
-      // Test baseline tracking
-      await service.beforeFileEditedCompat('/test/file.ts')
-      
-      // Test getting new diagnostics (should be empty since no IDE client is actually connected)
-      const newDiagnostics = await service.getNewDiagnosticsCompat()
-      expect(Array.isArray(newDiagnostics)).toBe(true)
-    })
-
-    test('should handle missing IDE client gracefully', async () => {
-      // Test with no connected clients
-      const noIdeClients = [
-        { type: 'disconnected', name: 'test-disconnected-2', config: {} } as unknown as MCPServerConnection,
-      ]
-      
-      await service.handleQueryStart(noIdeClients)
-      
-      // Should handle gracefully
-      const result = await service.getNewDiagnosticsCompat()
-      expect(result).toEqual([])
-    })
-  })
-})
--- a/src/services/diagnosticTracking.ts
+++ b/src/services/diagnosticTracking.ts
@@ -32,7 +32,7 @@ export class DiagnosticTrackingService {
  private baseline: Map<string, Diagnostic[]> = new Map()

  private initialized = false
-  private currentMcpClients: MCPServerConnection[] = []
+  private mcpClient: MCPServerConnection | undefined

  // Track when files were last processed/fetched
  private lastProcessedTimestamps: Map<string, number> = new Map()
@@ -48,17 +48,18 @@ export class DiagnosticTrackingService {
    return DiagnosticTrackingService.instance
  }

-  initialize() {
+  initialize(mcpClient: MCPServerConnection) {
    if (this.initialized) {
      return
    }

+    // TODO: Do not cache the connected mcpClient since it can change.
+    this.mcpClient = mcpClient
    this.initialized = true
  }

  async shutdown(): Promise<void> {
    this.initialized = false
-    this.currentMcpClients = []
    this.baseline.clear()
    this.rightFileDiagnosticsState.clear()
    this.lastProcessedTimestamps.clear()
@@ -74,46 +75,6 @@ export class DiagnosticTrackingService {
    this.lastProcessedTimestamps.clear()
  }

-  /**
-   * Get the current IDE client from stored MCP clients
-   */
-  private getCurrentIdeClient(): MCPServerConnection | undefined {
-    return getConnectedIdeClient(this.currentMcpClients)
-  }
-
-  /**
-   * Backward-compatible method that uses stored IDE client
-   */
-  async beforeFileEditedCompat(filePath: string): Promise<void> {
-    const ideClient = this.getCurrentIdeClient()
-    if (!ideClient) {
-      return
-    }
-    return await this.beforeFileEdited(filePath, ideClient)
-  }
-
-  /**
-   * Backward-compatible method that uses stored IDE client
-   */
-  async getNewDiagnosticsCompat(): Promise<DiagnosticFile[]> {
-    const ideClient = this.getCurrentIdeClient()
-    if (!ideClient) {
-      return []
-    }
-    return await this.getNewDiagnostics(ideClient)
-  }
-
-  /**
-   * Backward-compatible method that uses stored IDE client
-   */
-  async ensureFileOpenedCompat(fileUri: string): Promise<void> {
-    const ideClient = this.getCurrentIdeClient()
-    if (!ideClient) {
-      return
-    }
-    return await this.ensureFileOpened(fileUri, ideClient)
-  }
-
  private normalizeFileUri(fileUri: string): string {
    // Remove our protocol prefixes
    const protocolPrefixes = [
@@ -139,11 +100,11 @@ export class DiagnosticTrackingService {
   * Ensure a file is opened in the IDE before processing.
   * This is important for language services like diagnostics to work properly.
   */
-  async ensureFileOpened(fileUri: string, mcpClient: MCPServerConnection): Promise<void> {
+  async ensureFileOpened(fileUri: string): Promise<void> {
    if (
      !this.initialized ||
-      !mcpClient ||
-      mcpClient.type !== 'connected'
+      !this.mcpClient ||
+      this.mcpClient.type !== 'connected'
    ) {
      return
    }
@@ -160,7 +121,7 @@ export class DiagnosticTrackingService {
          selectToEndOfLine: false,
          makeFrontmost: false,
        },
-        mcpClient,
+        this.mcpClient,
      )
    } catch (error) {
      logError(error as Error)
@@ -171,11 +132,11 @@ export class DiagnosticTrackingService {
   * Capture baseline diagnostics for a specific file before editing.
   * This is called before editing a file to ensure we have a baseline to compare against.
   */
-  async beforeFileEdited(filePath: string, mcpClient: MCPServerConnection): Promise<void> {
+  async beforeFileEdited(filePath: string): Promise<void> {
    if (
      !this.initialized ||
-      !mcpClient ||
-      mcpClient.type !== 'connected'
+      !this.mcpClient ||
+      this.mcpClient.type !== 'connected'
    ) {
      return
    }
@@ -186,7 +147,7 @@ export class DiagnosticTrackingService {
      const result = await callIdeRpc(
        'getDiagnostics',
        { uri: `file://${filePath}` },
-        mcpClient,
+        this.mcpClient,
      )
      const diagnosticFile = this.parseDiagnosticResult(result)[0]
      if (diagnosticFile) {
@@ -224,11 +185,11 @@ export class DiagnosticTrackingService {
   * Get new diagnostics from file://, _claude_fs_right, and _claude_fs_ URIs that aren't in the baseline.
   * Only processes diagnostics for files that have been edited.
   */
-  async getNewDiagnostics(mcpClient: MCPServerConnection): Promise<DiagnosticFile[]> {
+  async getNewDiagnostics(): Promise<DiagnosticFile[]> {
    if (
      !this.initialized ||
-      !mcpClient ||
-      mcpClient.type !== 'connected'
+      !this.mcpClient ||
+      this.mcpClient.type !== 'connected'
    ) {
      return []
    }
@@ -239,7 +200,7 @@ export class DiagnosticTrackingService {
      const result = await callIdeRpc(
        'getDiagnostics',
        {}, // Empty params fetches all diagnostics
-        mcpClient,
+        this.mcpClient,
      )
      allDiagnosticFiles = this.parseDiagnosticResult(result)
    } catch (_error) {
@@ -367,16 +328,13 @@ export class DiagnosticTrackingService {
   * @param shouldQuery Whether a query is actually being made (not just a command)
   */
  async handleQueryStart(clients: MCPServerConnection[]): Promise<void> {
-    // Store the current MCP clients for later use
-    this.currentMcpClients = clients
-
    // Only proceed if we should query and have clients
    if (!this.initialized) {
      // Find the connected IDE client
      const connectedIdeClient = getConnectedIdeClient(clients)

      if (connectedIdeClient) {
-        this.initialize()
+        this.initialize(connectedIdeClient)
      }
    } else {
      // Reset diagnostic tracking for new query loops
--- a/src/services/github/deviceFlow.test.ts
+++ b/src/services/github/deviceFlow.test.ts
@@ -1,4 +1,4 @@
-import { afterEach, describe, expect, mock, test } from 'bun:test'
+import { afterEach, beforeEach, describe, expect, mock, test } from 'bun:test'

 import {
  DEFAULT_GITHUB_DEVICE_SCOPE,
@@ -12,15 +12,22 @@ async function importFreshModule() {
  return import(`./deviceFlow.ts?ts=${Date.now()}-${Math.random()}`)
 }

-afterEach(() => {
-  mock.restore()
-})
-
 describe('requestDeviceCode', () => {
+  const originalFetch = globalThis.fetch
+
+  beforeEach(() => {
+    mock.restore()
+    globalThis.fetch = originalFetch
+  })
+
+  afterEach(() => {
+    globalThis.fetch = originalFetch
+  })
+
  test('parses successful device code response', async () => {
    const { requestDeviceCode } = await importFreshModule()

-    const fetchImpl = mock(() =>
+    globalThis.fetch = mock(() =>
      Promise.resolve(
        new Response(
          JSON.stringify({
@@ -37,7 +44,7 @@ describe('requestDeviceCode', () => {

    const r = await requestDeviceCode({
      clientId: 'test-client',
-      fetchImpl,
+      fetchImpl: globalThis.fetch,
    })
    expect(r.device_code).toBe('abc')
    expect(r.user_code).toBe('ABCD-1234')
@@ -50,17 +57,17 @@ describe('requestDeviceCode', () => {
    const { requestDeviceCode, GitHubDeviceFlowError } =
      await importFreshModule()

-    const fetchImpl = mock(() =>
+    globalThis.fetch = mock(() =>
      Promise.resolve(new Response('bad', { status: 500 })),
    )
    await expect(
-      requestDeviceCode({ clientId: 'x', fetchImpl }),
+      requestDeviceCode({ clientId: 'x', fetchImpl: globalThis.fetch }),
    ).rejects.toThrow(GitHubDeviceFlowError)
  })

  test('uses OAuth-safe default scope', async () => {
    let capturedScope = ''
-    const fetchImpl = mock((_url: RequestInfo | URL, init?: RequestInit) => {
+    globalThis.fetch = mock((_url: RequestInfo | URL, init?: RequestInit) => {
      const body = init?.body
      if (body instanceof URLSearchParams) {
        capturedScope = body.get('scope') ?? ''
@@ -80,7 +87,7 @@ describe('requestDeviceCode', () => {
      )
    })

-    await requestDeviceCode({ clientId: 'test-client', fetchImpl })
+    await requestDeviceCode({ clientId: 'test-client', fetchImpl: globalThis.fetch })
    expect(capturedScope).toBe(DEFAULT_GITHUB_DEVICE_SCOPE)
    expect(capturedScope).toBe('read:user')
  })
@@ -89,7 +96,7 @@ describe('requestDeviceCode', () => {
    const scopesSeen: string[] = []
    let callCount = 0

-    const fetchImpl = mock((_url: RequestInfo | URL, init?: RequestInit) => {
+    globalThis.fetch = mock((_url: RequestInfo | URL, init?: RequestInit) => {
      const body = init?.body
      const scope =
        body instanceof URLSearchParams
@@ -125,7 +132,7 @@ describe('requestDeviceCode', () => {
    const result = await requestDeviceCode({
      clientId: 'test-client',
      scope: 'read:user,models:read',
-      fetchImpl,
+      fetchImpl: globalThis.fetch,
    })

    expect(result.device_code).toBe('abc')
@@ -135,11 +142,17 @@ describe('requestDeviceCode', () => {
 })

 describe('pollAccessToken', () => {
+  const originalFetch = globalThis.fetch
+
+  afterEach(() => {
+    globalThis.fetch = originalFetch
+  })
+
  test('returns token when GitHub responds with access_token immediately', async () => {
    const { pollAccessToken } = await importFreshModule()

    let calls = 0
-    const fetchImpl = mock(() => {
+    globalThis.fetch = mock(() => {
      calls++
      return Promise.resolve(
        new Response(JSON.stringify({ access_token: 'tok-xyz' }), {
@@ -150,7 +163,7 @@ describe('pollAccessToken', () => {

    const token = await pollAccessToken('dev-code', {
      clientId: 'cid',
-      fetchImpl,
+      fetchImpl: globalThis.fetch,
    })
    expect(token).toBe('tok-xyz')
    expect(calls).toBe(1)
@@ -159,7 +172,7 @@ describe('pollAccessToken', () => {
  test('throws on access_denied', async () => {
    const { pollAccessToken } = await importFreshModule()

-    const fetchImpl = mock(() =>
+    globalThis.fetch = mock(() =>
      Promise.resolve(
        new Response(JSON.stringify({ error: 'access_denied' }), {
          status: 200,
@@ -169,17 +182,23 @@ describe('pollAccessToken', () => {
    await expect(
      pollAccessToken('dc', {
        clientId: 'c',
-        fetchImpl,
+        fetchImpl: globalThis.fetch,
      }),
    ).rejects.toThrow(/denied/)
  })
 })

 describe('exchangeForCopilotToken', () => {
+  const originalFetch = globalThis.fetch
+
+  afterEach(() => {
+    globalThis.fetch = originalFetch
+  })
+
  test('parses successful Copilot token response', async () => {
    const { exchangeForCopilotToken } = await importFreshModule()

-    const fetchImpl = mock(() =>
+    globalThis.fetch = mock(() =>
      Promise.resolve(
        new Response(
          JSON.stringify({
@@ -195,7 +214,7 @@ describe('exchangeForCopilotToken', () => {
      ),
    )

-    const result = await exchangeForCopilotToken('oauth-token', fetchImpl)
+    const result = await exchangeForCopilotToken('oauth-token', globalThis.fetch)
    expect(result.token).toBe('copilot-token-xyz')
    expect(result.expires_at).toBe(1700000000)
    expect(result.refresh_in).toBe(3600)
@@ -206,24 +225,24 @@ describe('exchangeForCopilotToken', () => {
    const { exchangeForCopilotToken, GitHubDeviceFlowError } =
      await importFreshModule()

-    const fetchImpl = mock(() =>
+    globalThis.fetch = mock(() =>
      Promise.resolve(new Response('unauthorized', { status: 401 })),
    )
    await expect(
-      exchangeForCopilotToken('bad-token', fetchImpl),
+      exchangeForCopilotToken('bad-token', globalThis.fetch),
    ).rejects.toThrow(GitHubDeviceFlowError)
  })

  test('throws on malformed response', async () => {
    const { exchangeForCopilotToken } = await importFreshModule()

-    const fetchImpl = mock(() =>
+    globalThis.fetch = mock(() =>
      Promise.resolve(
        new Response(JSON.stringify({ invalid: 'data' }), { status: 200 }),
      ),
    )
    await expect(
-      exchangeForCopilotToken('oauth-token', fetchImpl),
+      exchangeForCopilotToken('oauth-token', globalThis.fetch),
    ).rejects.toThrow(/Malformed/)
  })
 })
--- a/src/tools.ts
+++ b/src/tools.ts
@@ -48,6 +48,7 @@ import { TodoWriteTool } from './tools/TodoWriteTool/TodoWriteTool.js'
 import { ExitPlanModeV2Tool } from './tools/ExitPlanModeTool/ExitPlanModeV2Tool.js'
 import { TestingPermissionTool } from './tools/testing/TestingPermissionTool.js'
 import { GrepTool } from './tools/GrepTool/GrepTool.js'
+import { RepoMapTool } from './tools/RepoMapTool/RepoMapTool.js'
 // Lazy require to break circular dependency: tools.ts -> TeamCreateTool/TeamDeleteTool -> ... -> tools.ts
 /* eslint-disable @typescript-eslint/no-require-imports */
 const getTeamCreateTool = () =>
@@ -188,6 +189,7 @@ export function getAllBaseTools(): Tools {
    // trick as ripgrep). When available, find/grep in Claude's shell are aliased
    // to these fast tools, so the dedicated Glob/Grep tools are unnecessary.
    ...(hasEmbeddedSearchTools() ? [] : [GlobTool, GrepTool]),
+    RepoMapTool,
    ExitPlanModeV2Tool,
    FileReadTool,
    FileEditTool,
--- a/src/tools/FileEditTool/FileEditTool.ts
+++ b/src/tools/FileEditTool/FileEditTool.ts
@@ -422,7 +422,7 @@ export const FileEditTool = buildTool({
      activateConditionalSkillsForPaths([absoluteFilePath], cwd)
    }

-    await diagnosticTracker.beforeFileEditedCompat(absoluteFilePath)
+    await diagnosticTracker.beforeFileEdited(absoluteFilePath)

    // Ensure parent directory exists before the atomic read-modify-write section.
    // These awaits must stay OUTSIDE the critical section below — a yield between
--- a/src/tools/FileWriteTool/FileWriteTool.ts
+++ b/src/tools/FileWriteTool/FileWriteTool.ts
@@ -244,7 +244,7 @@ export const FileWriteTool = buildTool({
    // Activate conditional skills whose path patterns match this file
    activateConditionalSkillsForPaths([fullFilePath], cwd)

-    await diagnosticTracker.beforeFileEditedCompat(fullFilePath)
+    await diagnosticTracker.beforeFileEdited(fullFilePath)

    // Ensure parent directory exists before the atomic read-modify-write section.
    // Must stay OUTSIDE the critical section below (a yield between the staleness
--- a/src/tools/MCPTool/MCPTool.ts
+++ b/src/tools/MCPTool/MCPTool.ts
@@ -1,8 +1,7 @@
-import { Ajv } from 'ajv'
 import { z } from 'zod/v4'
-import { buildTool, type ToolDef, type ValidationResult } from '../../Tool.js'
+import { buildTool, type ToolDef } from '../../Tool.js'
 import { lazySchema } from '../../utils/lazySchema.js'
-import type { PermissionResult } from '../../types/permissions.js'
+import type { PermissionResult } from '../../utils/permissions/PermissionResult.js'
 import { isOutputLineTruncated } from '../../utils/terminal.js'
 import { DESCRIPTION, PROMPT } from './prompt.js'
 import {
@@ -38,8 +37,6 @@ export type Output = z.infer<OutputSchema>
 // Re-export MCPProgress from centralized types to break import cycles
 export type { MCPProgress } from '../../types/tools.js'

-const ajv = new Ajv({ strict: false })
-
 export const MCPTool = buildTool({
  isMcp: true,
  // Overridden in mcpClient.ts with the real MCP tool name + args
@@ -75,27 +72,6 @@ export const MCPTool = buildTool({
      message: 'MCPTool requires permission.',
    }
  },
-  async validateInput(input, context): Promise<ValidationResult> {
-    if (this.inputJSONSchema) {
-      try {
-        const validate = ajv.compile(this.inputJSONSchema)
-        if (!validate(input)) {
-          return {
-            result: false,
-            message: ajv.errorsText(validate.errors),
-            errorCode: 400,
-          }
-        }
-      } catch (error) {
-        return {
-          result: false,
-          message: `Failed to compile JSON schema for validation: ${error}`,
-          errorCode: 500,
-        }
-      }
-    }
-    return { result: true }
-  },
  renderToolUseMessage,
  // Overridden in mcpClient.ts
  userFacingName: () => 'mcp',
@@ -124,4 +100,3 @@ export const MCPTool = buildTool({
    }
  },
 } satisfies ToolDef<InputSchema, Output>)
-
--- a/src/tools/RepoMapTool/RepoMapTool.test.ts
+++ b/src/tools/RepoMapTool/RepoMapTool.test.ts
@@ -0,0 +1,167 @@
+import { beforeAll, describe, expect, test } from 'bun:test'
+import { cpSync, mkdtempSync, rmSync } from 'fs'
+import { tmpdir } from 'os'
+import { join } from 'path'
+import { initParser } from '../../context/repoMap/parser.js'
+import { invalidateCache } from '../../context/repoMap/index.js'
+import { RepoMapTool } from './RepoMapTool.js'
+import { getToolUseSummary } from './UI.js'
+
+const FIXTURE_ROOT = join(
+  import.meta.dir,
+  '..',
+  '..',
+  'context',
+  'repoMap',
+  '__fixtures__',
+  'mini-repo',
+)
+const FIXTURE_FILES = [
+  'fileA.ts',
+  'fileB.ts',
+  'fileC.ts',
+  'fileD.ts',
+  'fileE.ts',
+]
+
+beforeAll(async () => {
+  await initParser()
+})
+
+
+describe('RepoMapTool schema', () => {
+  test('validates a minimal input {}', () => {
+    const schema = RepoMapTool.inputSchema
+    const result = schema.safeParse({})
+    expect(result.success).toBe(true)
+  })
+
+  test('rejects max_tokens below 256', () => {
+    const schema = RepoMapTool.inputSchema
+    const result = schema.safeParse({ max_tokens: 100 })
+    expect(result.success).toBe(false)
+  })
+
+  test('rejects max_tokens above 16384', () => {
+    const schema = RepoMapTool.inputSchema
+    const result = schema.safeParse({ max_tokens: 20000 })
+    expect(result.success).toBe(false)
+  })
+
+  test('accepts focus_files as string[]', () => {
+    const schema = RepoMapTool.inputSchema
+    const result = schema.safeParse({
+      focus_files: ['src/tools/', 'src/context.ts'],
+    })
+    expect(result.success).toBe(true)
+  })
+})
+
+describe('RepoMapTool call', () => {
+  test('returns a rendered map for a directory', async () => {
+    const tempDir = mkdtempSync(join(tmpdir(), 'repomap-tool-'))
+    try {
+      for (const f of FIXTURE_FILES) {
+        cpSync(join(FIXTURE_ROOT, f), join(tempDir, f))
+      }
+
+      // We need to call buildRepoMap directly since getCwd patching is complex
+      const { buildRepoMap } = await import(
+        '../../context/repoMap/index.js'
+      )
+      const result = await buildRepoMap({
+        root: tempDir,
+        maxTokens: 1024,
+      })
+
+      expect(result.map.length).toBeGreaterThan(0)
+      expect(result.fileCount).toBeGreaterThan(0)
+      expect(result.totalFileCount).toBe(5)
+      expect(result.tokenCount).toBeGreaterThan(0)
+      expect(result.tokenCount).toBeLessThanOrEqual(1024)
+    } finally {
+      rmSync(tempDir, { recursive: true, force: true })
+      invalidateCache(tempDir)
+    }
+  })
+
+  test('respects max_tokens parameter', async () => {
+    const tempDir = mkdtempSync(join(tmpdir(), 'repomap-tool-'))
+    try {
+      for (const f of FIXTURE_FILES) {
+        cpSync(join(FIXTURE_ROOT, f), join(tempDir, f))
+      }
+
+      const { buildRepoMap } = await import(
+        '../../context/repoMap/index.js'
+      )
+
+      const small = await buildRepoMap({ root: tempDir, maxTokens: 256 })
+      const large = await buildRepoMap({ root: tempDir, maxTokens: 4096 })
+
+      expect(small.tokenCount).toBeLessThanOrEqual(256)
+      // Large budget should include more or equal content
+      expect(large.map.length).toBeGreaterThanOrEqual(small.map.length)
+    } finally {
+      rmSync(tempDir, { recursive: true, force: true })
+      invalidateCache(tempDir)
+    }
+  })
+
+  test('focus_files boosts specified files in the ranking', async () => {
+    const tempDir = mkdtempSync(join(tmpdir(), 'repomap-tool-'))
+    try {
+      for (const f of FIXTURE_FILES) {
+        cpSync(join(FIXTURE_ROOT, f), join(tempDir, f))
+      }
+
+      const { buildRepoMap } = await import(
+        '../../context/repoMap/index.js'
+      )
+
+      // Without focus, fileE is ranked last (isolated)
+      const noFocus = await buildRepoMap({ root: tempDir, maxTokens: 2048 })
+      const lines = noFocus.map.split('\n')
+      const fileEPos = lines.findIndex(l => l === 'fileE.ts:')
+
+      // With focus on fileE
+      invalidateCache(tempDir)
+      const withFocus = await buildRepoMap({
+        root: tempDir,
+        maxTokens: 2048,
+        focusFiles: ['fileE.ts'],
+      })
+      const focusLines = withFocus.map.split('\n')
+      const fileEFocusPos = focusLines.findIndex(l => l === 'fileE.ts:')
+
+      // fileE should rank higher (earlier position) with focus
+      expect(fileEFocusPos).toBeLessThan(fileEPos)
+    } finally {
+      rmSync(tempDir, { recursive: true, force: true })
+      invalidateCache(tempDir)
+    }
+  })
+})
+
+describe('RepoMapTool properties', () => {
+  test('is marked read-only and concurrency-safe', () => {
+    expect(RepoMapTool.isReadOnly({})).toBe(true)
+    expect(RepoMapTool.isConcurrencySafe({})).toBe(true)
+  })
+})
+
+describe('RepoMapTool UI', () => {
+  test('getToolUseSummary returns descriptive string including focus', () => {
+    expect(getToolUseSummary(undefined)).toBe('Repository map')
+    expect(getToolUseSummary({})).toBe('Repository map')
+    expect(getToolUseSummary({ focus_files: ['src/tools/'] })).toContain(
+      'focus:',
+    )
+    expect(getToolUseSummary({ focus_files: ['src/tools/'] })).toContain(
+      'src/tools/',
+    )
+    expect(
+      getToolUseSummary({ focus_symbols: ['buildTool'] }),
+    ).toContain('buildTool')
+  })
+})
--- a/src/tools/RepoMapTool/RepoMapTool.ts
+++ b/src/tools/RepoMapTool/RepoMapTool.ts
@@ -0,0 +1,176 @@
+import { z } from 'zod/v4'
+import { buildTool, type ToolDef } from '../../Tool.js'
+import { getCwd } from '../../utils/cwd.js'
+import { lazySchema } from '../../utils/lazySchema.js'
+import { checkReadPermissionForTool } from '../../utils/permissions/filesystem.js'
+import type { PermissionDecision } from '../../utils/permissions/PermissionResult.js'
+import { buildRepoMap } from '../../context/repoMap/index.js'
+import { REPO_MAP_TOOL_NAME, getDescription } from './prompt.js'
+import {
+  getToolUseSummary,
+  renderToolResultMessage,
+  renderToolUseErrorMessage,
+  renderToolUseMessage,
+} from './UI.js'
+
+const inputSchema = lazySchema(() =>
+  z.strictObject({
+    max_tokens: z
+      .number()
+      .int()
+      .min(256)
+      .max(16384)
+      .optional()
+      .describe(
+        'Maximum token budget for the rendered map. Higher values include more files. Default: 1024.',
+      ),
+    focus_files: z
+      .array(z.string())
+      .optional()
+      .describe(
+        'Relative file or directory paths to boost in the ranking (e.g. ["src/tools/", "src/context.ts"]).',
+      ),
+    focus_symbols: z
+      .array(z.string())
+      .optional()
+      .describe(
+        'Symbol names to boost — files defining these symbols rank higher (e.g. ["buildTool", "ToolUseContext"]).',
+      ),
+  }),
+)
+type InputSchema = ReturnType<typeof inputSchema>
+
+const outputSchema = lazySchema(() =>
+  z.object({
+    rendered: z.string(),
+    token_count: z.number(),
+    file_count: z.number(),
+    total_file_count: z.number(),
+    cache_hit: z.boolean(),
+    build_time_ms: z.number(),
+  }),
+)
+type OutputSchema = ReturnType<typeof outputSchema>
+
+type Output = z.infer<OutputSchema>
+
+export const RepoMapTool = buildTool({
+  name: REPO_MAP_TOOL_NAME,
+  searchHint: 'structural map of repository files and symbols',
+  maxResultSizeChars: 50_000,
+  async description() {
+    return getDescription()
+  },
+  userFacingName() {
+    return 'Repository map'
+  },
+  getToolUseSummary,
+  getActivityDescription(input) {
+    if (input?.focus_files?.length) {
+      return `Building repository map (focus: ${input.focus_files.join(', ')})`
+    }
+    return 'Building repository map'
+  },
+  get inputSchema(): InputSchema {
+    return inputSchema()
+  },
+  get outputSchema(): OutputSchema {
+    return outputSchema()
+  },
+  isConcurrencySafe() {
+    return true
+  },
+  isReadOnly() {
+    return true
+  },
+  isSearchOrReadCommand() {
+    return { isSearch: false, isRead: true }
+  },
+  toAutoClassifierInput(input) {
+    const parts: string[] = ['repomap']
+    if (input.focus_files?.length) parts.push(`focus: ${input.focus_files.join(',')}`)
+    return parts.join(' ')
+  },
+  async checkPermissions(input, context): Promise<PermissionDecision> {
+    const appState = context.getAppState()
+    return checkReadPermissionForTool(
+      RepoMapTool,
+      input,
+      appState.toolPermissionContext,
+    )
+  },
+  async prompt() {
+    return getDescription()
+  },
+  renderToolUseMessage,
+  renderToolUseErrorMessage,
+  renderToolResultMessage,
+  extractSearchText({ rendered }) {
+    return rendered
+  },
+  mapToolResultToToolResultBlockParam(output, toolUseID) {
+    const summary = [
+      `Repository map: ${output.file_count} files ranked (${output.total_file_count} total), ${output.token_count} tokens`,
+      output.cache_hit ? '(cached)' : `(built in ${output.build_time_ms}ms)`,
+    ].join(' ')
+
+    return {
+      tool_use_id: toolUseID,
+      type: 'tool_result',
+      content: `${summary}\n\n${output.rendered}`,
+    }
+  },
+  async call(
+    { max_tokens = 1024, focus_files, focus_symbols },
+    { abortController },
+  ) {
+    const root = getCwd()
+
+    // Resolve focus_symbols to file paths by searching the tag cache
+    let resolvedFocusFiles = focus_files ?? []
+    if (focus_symbols?.length) {
+      // Import the symbol lookup dynamically to avoid circular deps at module load
+      const { getRepoFiles } = await import('../../context/repoMap/gitFiles.js')
+      const { extractTags } = await import('../../context/repoMap/symbolExtractor.js')
+      const { initParser } = await import('../../context/repoMap/parser.js')
+
+      await initParser()
+      const files = await getRepoFiles(root)
+      const symbolFiles: string[] = []
+      const symbolSet = new Set(focus_symbols)
+
+      // Scan files for matching symbol definitions
+      for (const file of files) {
+        if (abortController.signal.aborted) break
+        const tags = await extractTags(file, root)
+        if (tags) {
+          const hasMatch = tags.tags.some(
+            t => t.kind === 'def' && symbolSet.has(t.name),
+          )
+          if (hasMatch) {
+            symbolFiles.push(file)
+          }
+        }
+      }
+
+      resolvedFocusFiles = [...resolvedFocusFiles, ...symbolFiles]
+    }
+
+    const result = await buildRepoMap({
+      root,
+      maxTokens: max_tokens,
+      focusFiles: resolvedFocusFiles.length > 0 ? resolvedFocusFiles : undefined,
+    })
+
+    const output: Output = {
+      rendered: result.map,
+      token_count: result.tokenCount,
+      file_count: result.fileCount,
+      total_file_count: result.totalFileCount,
+      cache_hit: result.cacheHit,
+      build_time_ms: result.buildTimeMs,
+    }
+
+    return { data: output }
+  },
+} satisfies ToolDef<InputSchema, Output>)
--- a/src/tools/RepoMapTool/UI.tsx
+++ b/src/tools/RepoMapTool/UI.tsx
@@ -0,0 +1,96 @@
+import type { ToolResultBlockParam } from '@anthropic-ai/sdk/resources/index.mjs'
+import React from 'react'
+import { FallbackToolUseErrorMessage } from '../../components/FallbackToolUseErrorMessage.js'
+import { MessageResponse } from '../../components/MessageResponse.js'
+import { TOOL_SUMMARY_MAX_LENGTH } from '../../constants/toolLimits.js'
+import { Text } from '../../ink.js'
+import type { ToolProgressData } from '../../Tool.js'
+import type { ProgressMessage } from '../../types/message.js'
+import { truncate } from '../../utils/format.js'
+
+type Output = {
+  rendered: string
+  token_count: number
+  file_count: number
+  total_file_count: number
+  cache_hit: boolean
+  build_time_ms: number
+}
+
+export function getToolUseSummary(
+  input:
+    | Partial<{
+        max_tokens?: number
+        focus_files?: string[]
+        focus_symbols?: string[]
+      }>
+    | undefined,
+): string | null {
+  if (!input) return 'Repository map'
+  const parts: string[] = []
+  if (input.focus_files?.length) {
+    parts.push(input.focus_files.join(', '))
+  }
+  if (input.focus_symbols?.length) {
+    parts.push(input.focus_symbols.join(', '))
+  }
+  if (parts.length > 0) {
+    return truncate(`Repository map (focus: ${parts.join('; ')})`, TOOL_SUMMARY_MAX_LENGTH)
+  }
+  return 'Repository map'
+}
+
+export function renderToolUseMessage(
+  input: Partial<{
+    max_tokens?: number
+    focus_files?: string[]
+    focus_symbols?: string[]
+  }>,
+): React.ReactNode {
+  const parts: string[] = []
+  if (input.max_tokens) {
+    parts.push(`max_tokens: ${input.max_tokens}`)
+  }
+  if (input.focus_files?.length) {
+    parts.push(`focus: ${input.focus_files.join(', ')}`)
+  }
+  if (input.focus_symbols?.length) {
+    parts.push(`symbols: ${input.focus_symbols.join(', ')}`)
+  }
+  return parts.length > 0 ? parts.join(', ') : null
+}
+
+export function renderToolResultMessage(
+  output: Output,
+  _progressMessages: ProgressMessage<ToolProgressData>[],
+  { verbose }: { verbose: boolean },
+): React.ReactNode {
+  const summary = `${output.file_count} files ranked, ${output.token_count} tokens${output.cache_hit ? ' (cached)' : `, ${output.build_time_ms}ms`}`
+
+  if (verbose) {
+    return (
+      <MessageResponse>
+        <Text>
+          Built repository map: {summary}
+          {'\n'}
+          ({output.total_file_count} total files considered)
+        </Text>
+      </MessageResponse>
+    )
+  }
+
+  return (
+    <MessageResponse height={1}>
+      <Text>
+        Built repository map: {summary}
+      </Text>
+    </MessageResponse>
+  )
+}
+
+export function renderToolUseErrorMessage(
+  result: ToolResultBlockParam['content'],
+  { verbose }: { verbose: boolean },
+): React.ReactNode {
+  return <FallbackToolUseErrorMessage result={result} verbose={verbose} />
+}
--- a/src/tools/RepoMapTool/prompt.ts
+++ b/src/tools/RepoMapTool/prompt.ts
@@ -0,0 +1,31 @@
+export const REPO_MAP_TOOL_NAME = 'RepoMap'
+
+export function getDescription(): string {
+  return `Build a structural map of the repository showing ranked files and their key signatures (functions, classes, types, interfaces).
+
+## When to use
+- At the start of a session on an unfamiliar repository to understand the codebase architecture
+- Before cross-file refactors to identify which files are structurally connected
+- When searching for where a concept or feature lives across the codebase
+- When the user asks "how is this repo organized" or "what are the important files"
+
+## When NOT to use
+- To read the contents of a specific file — use Read instead
+- To search for exact text or patterns — use Grep instead
+- To find files by name or glob pattern — use Glob instead
+- When you already know which files to examine
+
+## How it works
+The tool parses every supported source file (TypeScript, JavaScript, Python) using tree-sitter, extracts symbol definitions and references, builds a cross-file reference graph weighted by symbol importance (IDF), and ranks files using PageRank. The output is a token-budgeted summary showing the highest-ranked files with their key signatures (function/class/type declarations).
+
+## Parameters
+- **max_tokens**: Controls how many files fit in the output. Use 1024 for a quick overview, 4096+ for comprehensive maps. Default: 1024.
+- **focus_files**: Pass relative paths (e.g. \`["src/tools/"]\`) to boost specific files and their neighbors in the ranking. Use when the user mentions specific directories or files.
+- **focus_symbols**: Pass symbol names (e.g. \`["buildTool", "ToolUseContext"]\`) to boost files that define those symbols. Use when the user asks about specific functions or types.
+
+## Important notes
+- The map shows **signatures only**, not function bodies. Use Read to see implementations.
+- Results are **auto-cached** on disk — repeat calls with the same parameters return instantly.
+- Files are ranked by structural importance: files imported by many others rank highest.
+`
+}
--- a/src/tools/WebSearchTool/WebSearchTool.ts
+++ b/src/tools/WebSearchTool/WebSearchTool.ts
@@ -9,7 +9,6 @@ import { z } from 'zod/v4'
 import { getFeatureValue_CACHED_MAY_BE_STALE } from '../../services/analytics/growthbook.js'
 import { queryModelWithStreaming } from '../../services/api/claude.js'
 import { collectCodexCompletedResponse } from '../../services/api/codexShim.js'
-import { fetchWithProxyRetry } from '../../services/api/fetchWithProxyRetry.js'
 import {
  resolveCodexApiCredentials,
  resolveProviderRequest,
@@ -315,7 +314,7 @@ async function runCodexWebSearch(
    body.reasoning = request.reasoning
  }

-  const response = await fetchWithProxyRetry(`${request.baseUrl}/responses`, {
+  const response = await fetch(`${request.baseUrl}/responses`, {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json',
--- a/src/utils/Cursor.ts
+++ b/src/utils/Cursor.ts
@@ -148,42 +148,6 @@ type Position = {
  column: number
 }

-export function maskTextWithVisibleEdges(
-  value: string,
-  mask: string,
-  visiblePrefix = 3,
-  visibleSuffix = 3,
-): string {
-  if (!mask || !value) return value
-
-  const graphemes = Array.from(getGraphemeSegmenter().segment(value))
-  const secretGraphemeCount = graphemes.filter(
-    ({ segment }) => segment !== '\n',
-  ).length
-  const visibleCount = visiblePrefix + visibleSuffix
-
-  if (secretGraphemeCount <= visibleCount) {
-    return graphemes
-      .map(({ segment }) => (segment === '\n' ? segment : mask))
-      .join('')
-  }
-
-  let secretIndex = 0
-  return graphemes
-    .map(({ segment }) => {
-      if (segment === '\n') return segment
-
-      const nextSegment =
-        secretIndex < visiblePrefix ||
-        secretIndex >= secretGraphemeCount - visibleSuffix
-          ? segment
-          : mask
-      secretIndex += 1
-      return nextSegment
-    })
-    .join('')
-}
-
 export class Cursor {
  readonly offset: number
  constructor(
@@ -244,12 +208,7 @@ export class Cursor {
    maxVisibleLines?: number,
  ) {
    const { line, column } = this.getPosition()
-    const allLines = mask
-      ? new MeasuredText(
-          maskTextWithVisibleEdges(this.text, mask),
-          this.measuredText.columns,
-        ).getWrappedText()
-      : this.measuredText.getWrappedText()
+    const allLines = this.measuredText.getWrappedText()

    const startLine = this.getViewportStartLine(maxVisibleLines)
    const endLine =
@@ -262,6 +221,23 @@ export class Cursor {
      .map((text, i) => {
        const currentLine = i + startLine
        let displayText = text
+        if (mask) {
+          const graphemes = Array.from(getGraphemeSegmenter().segment(text))
+          if (currentLine === allLines.length - 1) {
+            // Last line: mask all but the trailing 6 chars so the user can
+            // confirm they pasted the right thing without exposing the full token
+            const visibleCount = Math.min(6, graphemes.length)
+            const maskCount = graphemes.length - visibleCount
+            const splitOffset =
+              graphemes.length > visibleCount ? graphemes[maskCount]!.index : 0
+            displayText = mask.repeat(maskCount) + text.slice(splitOffset)
+          } else {
+            // Earlier wrapped lines: fully mask. Previously only the last line
+            // was masked, leaking the start of the token on narrow terminals
+            // where the pasted OAuth code wraps across multiple lines.
+            displayText = mask.repeat(graphemes.length)
+          }
+        }
        // looking for the line with the cursor
        if (line !== currentLine) return displayText.trimEnd()

--- a/src/utils/api.test.ts
+++ b/src/utils/api.test.ts
@@ -78,28 +78,3 @@ test('toolToAPISchema keeps skill required for SkillTool', async () => {
    required: ['skill'],
  })
 })
-
-test('toolToAPISchema removes extra required keys not in properties (MCP schema sanitization)', async () => {
-  const schema = await toolToAPISchema(
-    {
-      name: 'mcp__test__create_object',
-      inputSchema: z.strictObject({}),
-      inputJSONSchema: {
-        type: 'object',
-        properties: {
-          name: { type: 'string' },
-        },
-        required: ['name', 'attributes'],
-      },
-      prompt: async () => 'Create an object',
-    } as unknown as Tool,
-    {
-      getToolPermissionContext: async () => getEmptyToolPermissionContext(),
-      tools: [] as unknown as Tools,
-      agents: [],
-    },
-  )
-
-  const inputSchema = (schema as { input_schema: { required?: string[] } }).input_schema
-  expect(inputSchema.required).toEqual(['name'])
-})
--- a/src/utils/api.ts
+++ b/src/utils/api.ts
@@ -111,60 +111,11 @@ function filterSwarmFieldsFromSchema(
      delete filteredProps[field]
    }
    filtered.properties = filteredProps
-
-    // Keep `required` in sync after removing properties
-    if (Array.isArray(filtered.required)) {
-      filtered.required = filtered.required.filter(
-        (key: string) => key in filteredProps,
-      )
-    }
  }

  return filtered
 }

-/**
- * Ensure `required` only lists keys present in `properties`.
- * MCP servers may emit schemas where these are out of sync, causing
- * API 400 errors ("Extra required key supplied").
- * Recurses into nested object schemas.
- */
-function sanitizeSchemaRequired(
-  schema: Anthropic.Tool.InputSchema,
-): Anthropic.Tool.InputSchema {
-  if (!schema || typeof schema !== 'object') {
-    return schema
-  }
-
-  const result = { ...schema }
-  const props = result.properties as Record<string, unknown> | undefined
-
-  if (props && Array.isArray(result.required)) {
-    result.required = result.required.filter(
-      (key: string) => key in props,
-    )
-  }
-
-  // Recurse into nested object properties
-  if (props) {
-    const sanitizedProps = { ...props }
-    for (const [key, value] of Object.entries(sanitizedProps)) {
-      if (
-        value &&
-        typeof value === 'object' &&
-        (value as Record<string, unknown>).type === 'object'
-      ) {
-        sanitizedProps[key] = sanitizeSchemaRequired(
-          value as Anthropic.Tool.InputSchema,
-        )
-      }
-    }
-    result.properties = sanitizedProps
-  }
-
-  return result
-}
-
 export async function toolToAPISchema(
  tool: Tool,
  options: {
@@ -205,7 +156,7 @@ export async function toolToAPISchema(
    // Use tool's JSON schema directly if provided, otherwise convert Zod schema
    let input_schema = (
      'inputJSONSchema' in tool && tool.inputJSONSchema
-        ? sanitizeSchemaRequired(tool.inputJSONSchema as Anthropic.Tool.InputSchema)
+        ? tool.inputJSONSchema
        : zodToJsonSchema(tool.inputSchema)
    ) as Anthropic.Tool.InputSchema

--- a/src/utils/attachments.ts
+++ b/src/utils/attachments.ts
@@ -2882,7 +2882,7 @@ async function getDiagnosticAttachments(
  }

  // Get new diagnostics from the tracker (IDE diagnostics via MCP)
-  const newDiagnostics = await diagnosticTracker.getNewDiagnosticsCompat()
+  const newDiagnostics = await diagnosticTracker.getNewDiagnostics()
  if (newDiagnostics.length === 0) {
    return []
  }
--- a/src/utils/config.ts
+++ b/src/utils/config.ts
@@ -155,7 +155,7 @@ export {
  NOTIFICATION_CHANNELS,
 } from './configConstants.js'

-import type { EDITOR_MODES, NOTIFICATION_CHANNELS, PROVIDERS } from './configConstants.js'
+import type { EDITOR_MODES, NOTIFICATION_CHANNELS } from './configConstants.js'

 export type NotificationChannel = (typeof NOTIFICATION_CHANNELS)[number]

@@ -181,12 +181,10 @@ export type DiffTool = 'terminal' | 'auto'

 export type OutputStyle = string

-export type Providers = typeof PROVIDERS[number]
-
 export type ProviderProfile = {
  id: string
  name: string
-  provider: Providers
+  provider: 'openai' | 'anthropic'
  baseUrl: string
  model: string
  apiKey?: string
--- a/src/utils/configConstants.ts
+++ b/src/utils/configConstants.ts
@@ -19,5 +19,3 @@ export const EDITOR_MODES = ['normal', 'vim'] as const
 // 'in-process' = in-process teammates running in same process
 // 'auto' = automatically choose based on context (default)
 export const TEAMMATE_MODES = ['auto', 'tmux', 'in-process'] as const
-
-export const PROVIDERS = ['openai', 'anthropic', 'mistral', 'gemini'] as const
--- a/src/utils/context.test.ts
+++ b/src/utils/context.test.ts
@@ -9,7 +9,6 @@ import {
 const originalEnv = {
  CLAUDE_CODE_USE_OPENAI: process.env.CLAUDE_CODE_USE_OPENAI,
  CLAUDE_CODE_MAX_OUTPUT_TOKENS: process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS,
-  OPENAI_MODEL: process.env.OPENAI_MODEL,
 }

 afterEach(() => {
@@ -24,17 +23,11 @@ afterEach(() => {
    process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS =
      originalEnv.CLAUDE_CODE_MAX_OUTPUT_TOKENS
  }
-  if (originalEnv.OPENAI_MODEL === undefined) {
-    delete process.env.OPENAI_MODEL
-  } else {
-    process.env.OPENAI_MODEL = originalEnv.OPENAI_MODEL
-  }
 })

 test('deepseek-chat uses provider-specific context and output caps', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-  delete process.env.OPENAI_MODEL

  expect(getContextWindowForModel('deepseek-chat')).toBe(128_000)
  expect(getModelMaxOutputTokens('deepseek-chat')).toEqual({
@@ -47,7 +40,6 @@ test('deepseek-chat uses provider-specific context and output caps', () => {
 test('deepseek-chat clamps oversized max output overrides to the provider limit', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS = '32000'
-  delete process.env.OPENAI_MODEL

  expect(getMaxOutputTokensForModel('deepseek-chat')).toBe(8_192)
 })
@@ -55,7 +47,6 @@ test('deepseek-chat clamps oversized max output overrides to the provider limit'
 test('gpt-4o uses provider-specific context and output caps', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-  delete process.env.OPENAI_MODEL

  expect(getContextWindowForModel('gpt-4o')).toBe(128_000)
  expect(getModelMaxOutputTokens('gpt-4o')).toEqual({
@@ -68,7 +59,6 @@ test('gpt-4o uses provider-specific context and output caps', () => {
 test('gpt-4o clamps oversized max output overrides to the provider limit', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS = '32000'
-  delete process.env.OPENAI_MODEL

  expect(getMaxOutputTokensForModel('gpt-4o')).toBe(16_384)
 })
@@ -76,7 +66,6 @@ test('gpt-4o clamps oversized max output overrides to the provider limit', () =>
 test('gpt-5.4 family uses provider-specific context and output caps', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-  delete process.env.OPENAI_MODEL

  expect(getContextWindowForModel('gpt-5.4')).toBe(1_050_000)
  expect(getModelMaxOutputTokens('gpt-5.4')).toEqual({
@@ -109,7 +98,6 @@ test('gpt-5.4 family keeps large max output overrides within provider limits', (
 test('MiniMax-M2.7 uses explicit provider-specific context and output caps', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-  delete process.env.OPENAI_MODEL

  expect(getContextWindowForModel('MiniMax-M2.7')).toBe(204_800)
  expect(getModelMaxOutputTokens('MiniMax-M2.7')).toEqual({
@@ -122,7 +110,6 @@ test('MiniMax-M2.7 uses explicit provider-specific context and output caps', ()
 test('unknown openai-compatible models use the 128k fallback window (not 8k, see #635)', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-  delete process.env.OPENAI_MODEL

  expect(getContextWindowForModel('some-unknown-3p-model')).toBe(128_000)
 })
@@ -130,7 +117,6 @@ test('unknown openai-compatible models use the 128k fallback window (not 8k, see
 test('MiniMax-M2.5 and M2.1 use explicit provider-specific context and output caps', () => {
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-  delete process.env.OPENAI_MODEL

  expect(getContextWindowForModel('MiniMax-M2.5')).toBe(204_800)
  expect(getContextWindowForModel('MiniMax-M2.5-highspeed')).toBe(204_800)
@@ -141,116 +127,3 @@ test('MiniMax-M2.5 and M2.1 use explicit provider-specific context and output ca
    upperLimit: 131_072,
  })
 })
-
-test('DashScope qwen3.6-plus uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('qwen3.6-plus')).toBe(1_000_000)
-  expect(getModelMaxOutputTokens('qwen3.6-plus')).toEqual({
-    default: 65_536,
-    upperLimit: 65_536,
-  })
-  expect(getMaxOutputTokensForModel('qwen3.6-plus')).toBe(65_536)
-})
-
-test('DashScope qwen3.5-plus uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('qwen3.5-plus')).toBe(1_000_000)
-  expect(getModelMaxOutputTokens('qwen3.5-plus')).toEqual({
-    default: 65_536,
-    upperLimit: 65_536,
-  })
-  expect(getMaxOutputTokensForModel('qwen3.5-plus')).toBe(65_536)
-})
-
-test('DashScope qwen3-coder-plus uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('qwen3-coder-plus')).toBe(1_000_000)
-  expect(getModelMaxOutputTokens('qwen3-coder-plus')).toEqual({
-    default: 65_536,
-    upperLimit: 65_536,
-  })
-})
-
-test('DashScope qwen3-coder-next uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('qwen3-coder-next')).toBe(262_144)
-  expect(getModelMaxOutputTokens('qwen3-coder-next')).toEqual({
-    default: 65_536,
-    upperLimit: 65_536,
-  })
-})
-
-test('DashScope qwen3-max uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('qwen3-max')).toBe(262_144)
-  expect(getModelMaxOutputTokens('qwen3-max')).toEqual({
-    default: 32_768,
-    upperLimit: 32_768,
-  })
-})
-
-test('DashScope qwen3-max dated variant resolves to base entry via prefix match', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('qwen3-max-2026-01-23')).toBe(262_144)
-  expect(getModelMaxOutputTokens('qwen3-max-2026-01-23')).toEqual({
-    default: 32_768,
-    upperLimit: 32_768,
-  })
-})
-
-test('DashScope kimi-k2.5 uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('kimi-k2.5')).toBe(262_144)
-  expect(getModelMaxOutputTokens('kimi-k2.5')).toEqual({
-    default: 32_768,
-    upperLimit: 32_768,
-  })
-})
-
-test('DashScope glm-5 uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('glm-5')).toBe(202_752)
-  expect(getModelMaxOutputTokens('glm-5')).toEqual({
-    default: 16_384,
-    upperLimit: 16_384,
-  })
-})
-
-test('DashScope glm-4.7 uses provider-specific context and output caps', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  delete process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS
-
-  expect(getContextWindowForModel('glm-4.7')).toBe(202_752)
-  expect(getModelMaxOutputTokens('glm-4.7')).toEqual({
-    default: 16_384,
-    upperLimit: 16_384,
-  })
-})
-
-test('DashScope models clamp oversized max output overrides to the provider limit', () => {
-  process.env.CLAUDE_CODE_USE_OPENAI = '1'
-  process.env.CLAUDE_CODE_MAX_OUTPUT_TOKENS = '100000'
-
-  expect(getMaxOutputTokensForModel('qwen3.6-plus')).toBe(65_536)
-  expect(getMaxOutputTokensForModel('qwen3.5-plus')).toBe(65_536)
-  expect(getMaxOutputTokensForModel('qwen3-coder-next')).toBe(65_536)
-  expect(getMaxOutputTokensForModel('qwen3-max')).toBe(32_768)
-  expect(getMaxOutputTokensForModel('kimi-k2.5')).toBe(32_768)
-  expect(getMaxOutputTokensForModel('glm-5')).toBe(16_384)
-})
--- a/src/utils/model/configs.ts
+++ b/src/utils/model/configs.ts
@@ -37,8 +37,6 @@ export const CLAUDE_3_7_SONNET_CONFIG = {
  gemini: 'gemini-2.0-flash',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_3_5_V2_SONNET_CONFIG = {
@@ -50,8 +48,6 @@ export const CLAUDE_3_5_V2_SONNET_CONFIG = {
  gemini: 'gemini-2.0-flash',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_3_5_HAIKU_CONFIG = {
@@ -63,8 +59,6 @@ export const CLAUDE_3_5_HAIKU_CONFIG = {
  gemini: 'gemini-2.0-flash-lite',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_HAIKU_4_5_CONFIG = {
@@ -76,8 +70,6 @@ export const CLAUDE_HAIKU_4_5_CONFIG = {
  gemini: 'gemini-2.0-flash-lite',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_SONNET_4_CONFIG = {
@@ -89,8 +81,6 @@ export const CLAUDE_SONNET_4_CONFIG = {
  gemini: 'gemini-2.0-flash',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_SONNET_4_5_CONFIG = {
@@ -102,8 +92,6 @@ export const CLAUDE_SONNET_4_5_CONFIG = {
  gemini: 'gemini-2.0-flash',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_CONFIG = {
@@ -115,8 +103,6 @@ export const CLAUDE_OPUS_4_CONFIG = {
  gemini: 'gemini-2.5-pro-preview-03-25',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_1_CONFIG = {
@@ -128,8 +114,6 @@ export const CLAUDE_OPUS_4_1_CONFIG = {
  gemini: 'gemini-2.5-pro-preview-03-25',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_5_CONFIG = {
@@ -141,8 +125,6 @@ export const CLAUDE_OPUS_4_5_CONFIG = {
  gemini: 'gemini-2.5-pro-preview-03-25',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_6_CONFIG = {
@@ -154,8 +136,6 @@ export const CLAUDE_OPUS_4_6_CONFIG = {
  gemini: 'gemini-2.5-pro-preview-03-25',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 export const CLAUDE_SONNET_4_6_CONFIG = {
@@ -167,8 +147,6 @@ export const CLAUDE_SONNET_4_6_CONFIG = {
  gemini: 'gemini-2.0-flash',
  github: 'github:copilot',
  codex: 'gpt-5.4',
-  'nvidia-nim': 'nvidia/llama-3.1-nemotron-70b-instruct',
-  minimax: 'MiniMax-M2.5',
 } as const satisfies ModelConfig

 // @[MODEL LAUNCH]: Register the new config here.
--- a/src/utils/model/minimaxModels.ts
+++ b/src/utils/model/minimaxModels.ts
@@ -1,46 +0,0 @@
-/**
- * MiniMax model list for the /model picker.
- * Full model catalog from MiniMax API.
- */
-
-import type { ModelOption } from './modelOptions.js'
-import { getAPIProvider } from './providers.js'
-import { isEnvTruthy } from '../envUtils.js'
-
-export function isMiniMaxProvider(): boolean {
-  if (isEnvTruthy(process.env.MINIMAX_API_KEY)) {
-    return true
-  }
-  const baseUrl = process.env.OPENAI_BASE_URL ?? ''
-  if (baseUrl.includes('minimax')) {
-    return true
-  }
-  return getAPIProvider() === 'minimax'
-}
-
-function getMiniMaxModels(): ModelOption[] {
-  return [
-    // Latest Generation Models - use correct MiniMax naming with M prefix
-    { value: 'MiniMax-M2', label: 'MiniMax M2', description: 'MoE model - 131K context - Chat/Code/Reasoning' },
-    { value: 'MiniMax-M2.1', label: 'MiniMax M2.1', description: 'Enhanced - 200K context - Vision' },
-    { value: 'MiniMax-M2.5', label: 'MiniMax M2.5', description: 'Flagship - 256K context - Vision/Function-calling' },
-    { value: 'MiniMax-Text-01', label: 'MiniMax Text 01', description: 'Text-focused - 512K context - FREE' },
-    { value: 'MiniMax-Text-01-Preview', label: 'MiniMax Text 01 Preview', description: 'Preview - 256K context - FREE' },
-    { value: 'MiniMax-Vision-01', label: 'MiniMax Vision 01', description: 'Vision model - 32K context' },
-    { value: 'MiniMax-Vision-01-Fast', label: 'MiniMax Vision 01 Fast', description: 'Fast vision - 16K context - FREE' },
-    // Legacy free tier models
-    { value: 'abab6.5s-chat', label: 'ABAB 6.5S Chat', description: 'Legacy free - 16K context' },
-    { value: 'abab6.5-chat', label: 'ABAB 6.5 Chat', description: 'Legacy free - 32K context' },
-    { value: 'abab6.5g-chat', label: 'ABAB 6.5G Chat', description: 'Generation 6.5 - 32K context' },
-    { value: 'abab6-chat', label: 'ABAB 6 Chat', description: 'Legacy - 8K context' },
-  ]
-}
-
-let cachedMiniMaxOptions: ModelOption[] | null = null
-
-export function getCachedMiniMaxModelOptions(): ModelOption[] {
-  if (!cachedMiniMaxOptions) {
-    cachedMiniMaxOptions = getMiniMaxModels()
-  }
-  return cachedMiniMaxOptions
-}
--- a/src/utils/model/model.github.test.ts
+++ b/src/utils/model/model.github.test.ts
@@ -1,57 +0,0 @@
-import { afterEach, beforeEach, expect, test } from 'bun:test'
-
-import { saveGlobalConfig } from '../config.js'
-import { getDefaultMainLoopModelSetting, getUserSpecifiedModelSetting } from './model.js'
-
-const env = {
-  CLAUDE_CODE_USE_GITHUB: process.env.CLAUDE_CODE_USE_GITHUB,
-  CLAUDE_CODE_USE_OPENAI: process.env.CLAUDE_CODE_USE_OPENAI,
-  CLAUDE_CODE_USE_GEMINI: process.env.CLAUDE_CODE_USE_GEMINI,
-  CLAUDE_CODE_USE_BEDROCK: process.env.CLAUDE_CODE_USE_BEDROCK,
-  CLAUDE_CODE_USE_VERTEX: process.env.CLAUDE_CODE_USE_VERTEX,
-  CLAUDE_CODE_USE_FOUNDRY: process.env.CLAUDE_CODE_USE_FOUNDRY,
-  OPENAI_MODEL: process.env.OPENAI_MODEL,
-}
-
-beforeEach(() => {
-  process.env.CLAUDE_CODE_USE_GITHUB = '1'
-  delete process.env.CLAUDE_CODE_USE_OPENAI
-  delete process.env.CLAUDE_CODE_USE_GEMINI
-  delete process.env.CLAUDE_CODE_USE_BEDROCK
-  delete process.env.CLAUDE_CODE_USE_VERTEX
-  delete process.env.CLAUDE_CODE_USE_FOUNDRY
-  delete process.env.OPENAI_MODEL
-  saveGlobalConfig(current => ({
-    ...current,
-    model: ({ bad: true } as unknown) as string,
-  }))
-})
-
-afterEach(() => {
-  process.env.CLAUDE_CODE_USE_GITHUB = env.CLAUDE_CODE_USE_GITHUB
-  process.env.CLAUDE_CODE_USE_OPENAI = env.CLAUDE_CODE_USE_OPENAI
-  process.env.CLAUDE_CODE_USE_GEMINI = env.CLAUDE_CODE_USE_GEMINI
-  process.env.CLAUDE_CODE_USE_BEDROCK = env.CLAUDE_CODE_USE_BEDROCK
-  process.env.CLAUDE_CODE_USE_VERTEX = env.CLAUDE_CODE_USE_VERTEX
-  process.env.CLAUDE_CODE_USE_FOUNDRY = env.CLAUDE_CODE_USE_FOUNDRY
-  process.env.OPENAI_MODEL = env.OPENAI_MODEL
-  saveGlobalConfig(current => ({
-    ...current,
-    model: undefined,
-  }))
-})
-
-test('github default model setting ignores non-string saved model', () => {
-  const model = getDefaultMainLoopModelSetting()
-  expect(typeof model).toBe('string')
-  expect(model).not.toBe('[object Object]')
-  expect(model.length).toBeGreaterThan(0)
-})
-
-test('user specified model ignores non-string saved model', () => {
-  const model = getUserSpecifiedModelSetting()
-  if (model !== undefined && model !== null) {
-    expect(typeof model).toBe('string')
-    expect(model).not.toBe('[object Object]')
-  }
-})
--- a/src/utils/model/model.ts
+++ b/src/utils/model/model.ts
@@ -33,12 +33,6 @@ export type ModelShortName = string
 export type ModelName = string
 export type ModelSetting = ModelName | ModelAlias | null

-function normalizeModelSetting(value: unknown): ModelName | ModelAlias | undefined {
-  if (typeof value !== 'string') return undefined
-  const trimmed = value.trim()
-  return trimmed.length > 0 ? trimmed : undefined
-}
-
 export function getSmallFastModel(): ModelName {
  if (process.env.ANTHROPIC_SMALL_FAST_MODEL) return process.env.ANTHROPIC_SMALL_FAST_MODEL
  // For Gemini provider, use a fast model
@@ -88,7 +82,6 @@ export function getUserSpecifiedModelSetting(): ModelSetting | undefined {
    specifiedModel = modelOverride
  } else {
    const settings = getSettings_DEPRECATED() || {}
-    const setting = normalizeModelSetting(settings.model)
    // Read the model env var that matches the active provider to prevent
    // cross-provider leaks (e.g. ANTHROPIC_MODEL sent to the OpenAI API).
    const provider = getAPIProvider()
@@ -97,7 +90,7 @@ export function getUserSpecifiedModelSetting(): ModelSetting | undefined {
      (provider === 'mistral' ? process.env.MISTRAL_MODEL : undefined) ||
      (provider === 'openai' || provider === 'gemini' || provider === 'mistral' || provider === 'github' ? process.env.OPENAI_MODEL : undefined) ||
      (provider === 'firstParty' ? process.env.ANTHROPIC_MODEL : undefined) ||
-      setting ||
+      settings.model ||
      undefined
  }

@@ -271,11 +264,7 @@ export function getDefaultMainLoopModelSetting(): ModelName | ModelAlias {
  // GitHub Copilot provider: check settings.model first, then env, then default
  if (getAPIProvider() === 'github') {
    const settings = getSettings_DEPRECATED() || {}
-    return (
-      normalizeModelSetting(settings.model) ||
-      normalizeModelSetting(process.env.OPENAI_MODEL) ||
-      'github:copilot'
-    )
+    return settings.model || process.env.OPENAI_MODEL || 'github:copilot'
  }
  // Gemini provider: always use the configured Gemini model
  if (getAPIProvider() === 'gemini') {
@@ -606,10 +595,7 @@ export function getPublicModelName(model: ModelName): string {
 export function parseUserSpecifiedModel(
  modelInput: ModelName | ModelAlias,
 ): ModelName {
-  const modelInputTrimmed = normalizeModelSetting(modelInput)
-  if (!modelInputTrimmed) {
-    return getDefaultSonnetModel()
-  }
+  const modelInputTrimmed = modelInput.trim()
  const normalizedModel = modelInputTrimmed.toLowerCase()

  const has1mTag = has1mContext(normalizedModel)
--- a/src/utils/model/modelOptions.ts
+++ b/src/utils/model/modelOptions.ts
@@ -33,14 +33,8 @@ import {
 } from './model.js'
 import { has1mContext } from '../context.js'
 import { getGlobalConfig } from '../config.js'
-import {
-  getActiveOpenAIModelOptionsCache,
-  getActiveProviderProfile,
-  getProfileModelOptions,
-} from '../providerProfiles.js'
+import { getActiveOpenAIModelOptionsCache } from '../providerProfiles.js'
 import { getCachedOllamaModelOptions, isOllamaProvider } from './ollamaModels.js'
-import { getCachedNvidiaNimModelOptions, isNvidiaNimProvider } from './nvidiaNimModels.js'
-import { getCachedMiniMaxModelOptions, isMiniMaxProvider } from './minimaxModels.js'
 import { getAntModels } from './antModels.js'

 // @[MODEL LAUNCH]: Update all the available and default model option strings below.
@@ -396,26 +390,6 @@ function getModelOptionsBase(fastMode = false): ModelOption[] {
    return [defaultOption]
  }

-  // When using NVIDIA NIM, show models from the NVIDIA catalog
-  if (isNvidiaNimProvider()) {
-    const defaultOption = getDefaultOptionForUser(fastMode)
-    const nvidiaModels = getCachedNvidiaNimModelOptions()
-    if (nvidiaModels.length > 0) {
-      return [defaultOption, ...nvidiaModels]
-    }
-    return [defaultOption]
-  }
-
-  // When using MiniMax, show models from the MiniMax catalog
-  if (isMiniMaxProvider()) {
-    const defaultOption = getDefaultOptionForUser(fastMode)
-    const minimaxModels = getCachedMiniMaxModelOptions()
-    if (minimaxModels.length > 0) {
-      return [defaultOption, ...minimaxModels]
-    }
-    return [defaultOption]
-  }
-
  if (process.env.USER_TYPE === 'ant') {
    // Build options from antModels config
    const antModelOptions: ModelOption[] = getAntModels().map(m => ({
@@ -480,20 +454,6 @@ function getModelOptionsBase(fastMode = false): ModelOption[] {
    ]
  }

-  // When a provider profile's env is applied, collect its models so they
-  // can be appended to the standard picker options below.
-  // We check PROFILE_ENV_APPLIED to avoid the ?? profiles[0] fallback in
-  // getActiveProviderProfile which would affect users with inactive profiles.
-  const profileEnvApplied = process.env.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED === '1'
-  const profileModelOptions: ModelOption[] = []
-  if (profileEnvApplied) {
-    const activeProfile = getActiveProviderProfile()
-    if (activeProfile) {
-      const models = getProfileModelOptions(activeProfile)
-      profileModelOptions.push(...models)
-    }
-  }
-
  // PAYG 1P API: Default (Sonnet) + Sonnet 1M + Opus 4.6 + Opus 1M + Haiku
  if (getAPIProvider() === 'firstParty') {
    const payg1POptions = [getDefaultOptionForUser(fastMode)]
@@ -509,7 +469,6 @@ function getModelOptionsBase(fastMode = false): ModelOption[] {
      }
    }
    payg1POptions.push(getHaiku45Option())
-    payg1POptions.push(...profileModelOptions)
    return payg1POptions
  }

@@ -549,7 +508,6 @@ function getModelOptionsBase(fastMode = false): ModelOption[] {
  } else {
    payg3pOptions.push(getHaikuOption())
  }
-  payg3pOptions.push(...profileModelOptions)
  return payg3pOptions
 }

--- a/src/utils/model/nvidiaNimModels.ts
+++ b/src/utils/model/nvidiaNimModels.ts
@@ -1,161 +0,0 @@
-/**
- * NVIDIA NIM model list for the /model picker.
- * Filtered to chat/instruct models only - embedding, reward, safety, vision, etc. excluded.
- */
-
-import type { ModelOption } from './modelOptions.js'
-import { getAPIProvider } from './providers.js'
-import { isEnvTruthy } from '../envUtils.js'
-
-export function isNvidiaNimProvider(): boolean {
-  // Check if explicitly set via NVIDIA_NIM or via provider flag
-  if (isEnvTruthy(process.env.NVIDIA_NIM)) {
-    return true
-  }
-  // Also check if using NVIDIA NIM endpoint
-  const baseUrl = process.env.OPENAI_BASE_URL ?? ''
-  if (baseUrl.includes('nvidia') || baseUrl.includes('integrate.api.nvidia')) {
-    return true
-  }
-  return getAPIProvider() === 'nvidia-nim'
-}
-
-function getNvidiaNimModels(): ModelOption[] {
-  return [
-    // AGENTIC REASONING MODELS
-    { value: 'nvidia/cosmos-reason2-8b', label: 'Cosmos Reason 2 8B', description: 'Reasoning' },
-    { value: 'microsoft/phi-4-mini-flash-reasoning', label: 'Phi 4 Mini Flash Reasoning', description: 'Reasoning' },
-    { value: 'qwen/qwen3-next-80b-a3b-thinking', label: 'Qwen 3 Next 80B Thinking', description: 'Reasoning' },
-    { value: 'deepseek-ai/deepseek-r1-distill-qwen-32b', label: 'DeepSeek R1 Qwen 32B', description: 'Reasoning' },
-    { value: 'deepseek-ai/deepseek-r1-distill-qwen-14b', label: 'DeepSeek R1 Qwen 14B', description: 'Reasoning' },
-    { value: 'deepseek-ai/deepseek-r1-distill-qwen-7b', label: 'DeepSeek R1 Qwen 7B', description: 'Reasoning' },
-    { value: 'deepseek-ai/deepseek-r1-distill-llama-8b', label: 'DeepSeek R1 Llama 8B', description: 'Reasoning' },
-    { value: 'qwen/qwq-32b', label: 'QwQ 32B Reasoning', description: 'Reasoning' },
-    // CODE MODELS
-    { value: 'meta/codellama-70b', label: 'CodeLlama 70B', description: 'Code' },
-    { value: 'bigcode/starcoder2-15b', label: 'StarCoder2 15B', description: 'Code' },
-    { value: 'bigcode/starcoder2-7b', label: 'StarCoder2 7B', description: 'Code' },
-    { value: 'mistralai/codestral-22b-instruct-v0.1', label: 'Codestral 22B', description: 'Code' },
-    { value: 'mistralai/mamba-codestral-7b-v0.1', label: 'Mamba Codestral 7B', description: 'Code' },
-    { value: 'deepseek-ai/deepseek-coder-6.7b-instruct', label: 'DeepSeek Coder 6.7B', description: 'Code' },
-    { value: 'google/codegemma-7b', label: 'CodeGemma 7B', description: 'Code' },
-    { value: 'google/codegemma-1.1-7b', label: 'CodeGemma 1.1 7B', description: 'Code' },
-    { value: 'qwen/qwen2.5-coder-32b-instruct', label: 'Qwen 2.5 Coder 32B', description: 'Code' },
-    { value: 'qwen/qwen2.5-coder-7b-instruct', label: 'Qwen 2.5 Coder 7B', description: 'Code' },
-    { value: 'qwen/qwen3-coder-480b-a35b-instruct', label: 'Qwen 3 Coder 480B', description: 'Code' },
-    { value: 'ibm/granite-34b-code-instruct', label: 'Granite 34B Code', description: 'Code' },
-    { value: 'ibm/granite-8b-code-instruct', label: 'Granite 8B Code', description: 'Code' },
-    // NEMOTRON MODELS - NVIDIA Flagship
-    { value: 'nvidia/llama-3.1-nemotron-70b-instruct', label: 'Nemotron 70B Instruct', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/llama-3.1-nemotron-51b-instruct', label: 'Nemotron 51B Instruct', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/llama-3.1-nemotron-ultra-253b-v1', label: 'Nemotron Ultra 253B', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/llama-3.3-nemotron-super-49b-v1', label: 'Nemotron Super 49B v1', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/llama-3.3-nemotron-super-49b-v1.5', label: 'Nemotron Super 49B v1.5', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/nemotron-4-340b-instruct', label: 'Nemotron 4 340B', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/nemotron-3-super-120b-a12b', label: 'Nemotron 3 Super 120B', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/nemotron-3-nano-30b-a3b', label: 'Nemotron 3 Nano 30B', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/nemotron-mini-4b-instruct', label: 'Nemotron Mini 4B', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/llama-3.1-nemotron-nano-8b-v1', label: 'Nemotron Nano 8B', description: 'NVIDIA Flagship' },
-    { value: 'nvidia/llama-3.1-nemotron-nano-4b-v1.1', label: 'Nemotron Nano 4B v1.1', description: 'NVIDIA Flagship' },
-    // CHATQA MODELS
-    { value: 'nvidia/llama3-chatqa-1.5-70b', label: 'Llama3 ChatQA 1.5 70B', description: 'Chat' },
-    { value: 'nvidia/llama3-chatqa-1.5-8b', label: 'Llama3 ChatQA 1.5 8B', description: 'Chat' },
-    // META LLAMA MODELS
-    { value: 'meta/llama-3.1-405b-instruct', label: 'Llama 3.1 405B', description: 'Meta Llama' },
-    { value: 'meta/llama-3.1-70b-instruct', label: 'Llama 3.1 70B', description: 'Meta Llama' },
-    { value: 'meta/llama-3.1-8b-instruct', label: 'Llama 3.1 8B', description: 'Meta Llama' },
-    { value: 'meta/llama-3.2-90b-vision-instruct', label: 'Llama 3.2 90B Vision', description: 'Meta Llama' },
-    { value: 'meta/llama-3.2-11b-vision-instruct', label: 'Llama 3.2 11B Vision', description: 'Meta Llama' },
-    { value: 'meta/llama-3.2-3b-instruct', label: 'Llama 3.2 3B', description: 'Meta Llama' },
-    { value: 'meta/llama-3.2-1b-instruct', label: 'Llama 3.2 1B', description: 'Meta Llama' },
-    { value: 'meta/llama-3.3-70b-instruct', label: 'Llama 3.3 70B', description: 'Meta Llama' },
-    { value: 'meta/llama-4-maverick-17b-128e-instruct', label: 'Llama 4 Maverick 17B', description: 'Meta Llama' },
-    { value: 'meta/llama-4-scout-17b-16e-instruct', label: 'Llama 4 Scout 17B', description: 'Meta Llama' },
-    // GOOGLE GEMMA MODELS (text only - no vision)
-    { value: 'google/gemma-4-31b-it', label: 'Gemma 4 31B', description: 'Google Gemma' },
-    { value: 'google/gemma-3-27b-it', label: 'Gemma 3 27B', description: 'Google Gemma' },
-    { value: 'google/gemma-3-12b-it', label: 'Gemma 3 12B', description: 'Google Gemma' },
-    { value: 'google/gemma-3-4b-it', label: 'Gemma 3 4B', description: 'Google Gemma' },
-    { value: 'google/gemma-3-1b-it', label: 'Gemma 3 1B', description: 'Google Gemma' },
-    { value: 'google/gemma-3n-e4b-it', label: 'Gemma 3N E4B', description: 'Google Gemma' },
-    { value: 'google/gemma-3n-e2b-it', label: 'Gemma 3N E2B', description: 'Google Gemma' },
-    { value: 'google/gemma-2-27b-it', label: 'Gemma 2 27B', description: 'Google Gemma' },
-    { value: 'google/gemma-2-9b-it', label: 'Gemma 2 9B', description: 'Google Gemma' },
-    { value: 'google/gemma-2-2b-it', label: 'Gemma 2 2B', description: 'Google Gemma' },
-    // MISTRAL MODELS
-    { value: 'mistralai/mistral-large-3-675b-instruct-2512', label: 'Mistral Large 3 675B', description: 'Mistral' },
-    { value: 'mistralai/mistral-large-2-instruct', label: 'Mistral Large 2', description: 'Mistral' },
-    { value: 'mistralai/mistral-large', label: 'Mistral Large', description: 'Mistral' },
-    { value: 'mistralai/mistral-medium-3-instruct', label: 'Mistral Medium 3', description: 'Mistral' },
-    { value: 'mistralai/mistral-small-4-119b-2603', label: 'Mistral Small 4 119B', description: 'Mistral' },
-    { value: 'mistralai/mistral-small-3.1-24b-instruct-2503', label: 'Mistral Small 3.1 24B', description: 'Mistral' },
-    { value: 'mistralai/mistral-small-24b-instruct', label: 'Mistral Small 24B', description: 'Mistral' },
-    { value: 'mistralai/mistral-7b-instruct-v0.3', label: 'Mistral 7B v0.3', description: 'Mistral' },
-    { value: 'mistralai/mistral-7b-instruct-v0.2', label: 'Mistral 7B v0.2', description: 'Mistral' },
-    { value: 'mistralai/mixtral-8x22b-instruct-v0.1', label: 'Mixtral 8x22B', description: 'Mistral' },
-    { value: 'mistralai/mixtral-8x22b-instruct-v0.1', label: 'Mixtral 8x22B Instruct', description: 'Mistral' },
-    { value: 'mistralai/mixtral-8x7b-instruct-v0.1', label: 'Mixtral 8x7B', description: 'Mistral' },
-    { value: 'mistralai/mistral-nemotron', label: 'Mistral Nemotron', description: 'Mistral' },
-    { value: 'mistralai/mathstral-7b-v0.1', label: 'Mathstral 7B', description: 'Math' },
-    { value: 'mistralai/ministral-14b-instruct-2512', label: 'Ministral 14B', description: 'Mistral' },
-    { value: 'mistralai/devstral-2-123b-instruct-2512', label: 'Devstral 2 123B', description: 'Code' },
-    { value: 'mistralai/magistral-small-2506', label: 'Magistral Small', description: 'Mistral' },
-    // MICROSOFT PHI MODELS (text only - no vision)
-    { value: 'microsoft/phi-4-multimodal-instruct', label: 'Phi 4 Multimodal', description: 'Multimodal' },
-    { value: 'microsoft/phi-4-mini-instruct', label: 'Phi 4 Mini', description: 'Phi' },
-    { value: 'microsoft/phi-3.5-mini-instruct', label: 'Phi 3.5 Mini', description: 'Phi' },
-    { value: 'microsoft/phi-3-small-128k-instruct', label: 'Phi 3 Small 128K', description: 'Phi' },
-    { value: 'microsoft/phi-3-small-8k-instruct', label: 'Phi 3 Small 8K', description: 'Phi' },
-    { value: 'microsoft/phi-3-medium-128k-instruct', label: 'Phi 3 Medium 128K', description: 'Phi' },
-    { value: 'microsoft/phi-3-medium-4k-instruct', label: 'Phi 3 Medium 4K', description: 'Phi' },
-    { value: 'microsoft/phi-3-mini-128k-instruct', label: 'Phi 3 Mini 128K', description: 'Phi' },
-    { value: 'microsoft/phi-3-mini-4k-instruct', label: 'Phi 3 Mini 4K', description: 'Phi' },
-    // QWEN MODELS
-    { value: 'qwen/qwen3.5-397b-a17b', label: 'Qwen 3.5 397B', description: 'Qwen' },
-    { value: 'qwen/qwen3.5-122b-a10b', label: 'Qwen 3.5 122B', description: 'Qwen' },
-    { value: 'qwen/qwen3-next-80b-a3b-instruct', label: 'Qwen 3 Next 80B', description: 'Qwen' },
-    { value: 'qwen/qwen2.5-7b-instruct', label: 'Qwen 2.5 7B', description: 'Qwen' },
-    { value: 'qwen/qwen2-7b-instruct', label: 'Qwen 2 7B', description: 'Qwen' },
-    { value: 'qwen/qwen3-32b', label: 'Qwen 3 32B', description: 'Qwen' },
-    { value: 'qwen/qwen3-8b', label: 'Qwen 3 8B', description: 'Qwen' },
-    // DEEPSEEK MODELS
-    { value: 'deepseek-ai/deepseek-r1', label: 'DeepSeek R1', description: 'DeepSeek' },
-    { value: 'deepseek-ai/deepseek-v3', label: 'DeepSeek V3', description: 'DeepSeek' },
-    { value: 'deepseek-ai/deepseek-v3.2', label: 'DeepSeek V3.2', description: 'DeepSeek' },
-    { value: 'deepseek-ai/deepseek-v3.1-terminus', label: 'DeepSeek V3.1 Terminus', description: 'DeepSeek' },
-    { value: 'deepseek-ai/deepseek-v3.1', label: 'DeepSeek V3.1', description: 'DeepSeek' },
-    // IBM GRANITE MODELS
-    { value: 'ibm/granite-3.3-8b-instruct', label: 'Granite 3.3 8B', description: 'IBM Granite' },
-    { value: 'ibm/granite-3.0-8b-instruct', label: 'Granite 3.0 8B', description: 'IBM Granite' },
-    { value: 'ibm/granite-3.0-3b-a800m-instruct', label: 'Granite 3.0 3B', description: 'IBM Granite' },
-    // OTHER MODELS
-    { value: 'databricks/dbrx-instruct', label: 'DBRX Instruct', description: 'Other' },
-    { value: '01-ai/yi-large', label: 'Yi Large', description: 'Other' },
-    { value: 'ai21labs/jamba-1.5-large-instruct', label: 'Jamba 1.5 Large', description: 'Other' },
-    { value: 'ai21labs/jamba-1.5-mini-instruct', label: 'Jamba 1.5 Mini', description: 'Other' },
-    { value: 'writer/palmyra-creative-122b', label: 'Palmyra Creative 122B', description: 'Other' },
-    { value: 'writer/palmyra-fin-70b-32k', label: 'Palmyra Fin 70B 32K', description: 'Other' },
-    { value: 'writer/palmyra-med-70b', label: 'Palmyra Med 70B', description: 'Other' },
-    { value: 'writer/palmyra-med-70b-32k', label: 'Palmyra Med 70B 32K', description: 'Other' },
-    // Z-AI GLM MODELS
-    { value: 'z-ai/glm5', label: 'GLM-5', description: 'Z-AI' },
-    { value: 'z-ai/glm4.7', label: 'GLM-4.7', description: 'Z-AI' },
-    // MINIMAX MODELS
-    { value: 'minimaxai/minimax-m2.5', label: 'MiniMax M2.5', description: 'MiniMax' },
-    // MOONSHOT KIMI MODELS
-    { value: 'moonshotai/kimi-k2.5', label: 'Kimi K2.5', description: 'Moonshot' },
-    { value: 'moonshotai/kimi-k2-instruct', label: 'Kimi K2 Instruct', description: 'Moonshot' },
-    { value: 'moonshotai/kimi-k2-thinking', label: 'Kimi K2 Thinking', description: 'Moonshot' },
-    { value: 'moonshotai/kimi-k2.5-thinking', label: 'Kimi K2.5 Thinking', description: 'Moonshot' },
-    { value: 'moonshotai/kimi-k2-instruct-0905', label: 'Kimi K2 Instruct 0905', description: 'Moonshot' },
-  ]
-}
-
-let cachedNvidiaNimOptions: ModelOption[] | null = null
-
-export function getCachedNvidiaNimModelOptions(): ModelOption[] {
-  if (!cachedNvidiaNimOptions) {
-    cachedNvidiaNimOptions = getNvidiaNimModels()
-  }
-  return cachedNvidiaNimOptions
-}
--- a/src/utils/model/openaiContextWindows.ts
+++ b/src/utils/model/openaiContextWindows.ts
@@ -104,57 +104,6 @@ const OPENAI_CONTEXT_WINDOWS: Record<string, number> = {
  'devstral-latest':          256_000,
  'ministral-3b-latest':      256_000,

-  // NVIDIA NIM - popular models
-  'nvidia/llama-3.1-nemotron-70b-instruct': 128_000,
-  'nvidia/llama-3.1-nemotron-ultra-253b-v1': 128_000,
-  'nvidia/nemotron-mini-4b-instruct': 32_768,
-  'meta/llama-3.1-405b-instruct': 128_000,
-  'meta/llama-3.1-70b-instruct': 128_000,
-  'meta/llama-3.1-8b-instruct': 128_000,
-  'meta/llama-3.2-90b-instruct': 128_000,
-  'meta/llama-3.2-1b-instruct': 128_000,
-  'meta/llama-3.2-3b-instruct': 128_000,
-  'meta/llama-3.3-70b-instruct': 128_000,
-  // Google Gemma via NVIDIA NIM
-  'google/gemma-2-27b-it': 8_192,
-  'google/gemma-2-9b-it': 8_192,
-  'google/gemma-3-27b-it': 131_072,
-  'google/gemma-3-12b-it': 131_072,
-  'google/gemma-3-4b-it': 131_072,
-  // DeepSeek via NVIDIA NIM
-  'deepseek-ai/deepseek-r1': 128_000,
-  'deepseek-ai/deepseek-v3': 128_000,
-  'deepseek-ai/deepseek-v3.2': 128_000,
-  // Qwen via NVIDIA NIM
-  'qwen/qwen3-32b': 128_000,
-  'qwen/qwen3-8b': 128_000,
-  'qwen/qwen2.5-7b-instruct': 32_768,
-  // Mistral via NVIDIA NIM
-  'mistralai/mistral-large-3-675b-instruct-2512': 256_000,
-  'mistralai/mistral-large-2-instruct': 256_000,
-  'mistralai/mistral-small-3.1-24b-instruct-2503': 32_768,
-  'mistralai/mixtral-8x7b-instruct-v0.1': 32_768,
-  // Microsoft Phi via NVIDIA NIM
-  'microsoft/phi-4-mini-instruct': 16_384,
-  'microsoft/phi-3.5-mini-instruct': 16_384,
-  'microsoft/phi-3-mini-128k-instruct': 128_000,
-  // IBM Granite via NVIDIA NIM
-  'ibm/granite-3.3-8b-instruct': 8_192,
-  'ibm/granite-8b-code-instruct': 8_192,
-  // GLM models via NVIDIA NIM
-  'z-ai/glm5': 200_000,
-  'z-ai/glm4.7': 128_000,
-  // Kimi models via NVIDIA NIM
-  'moonshotai/kimi-k2.5': 200_000,
-  'moonshotai/kimi-k2-instruct': 128_000,
-  // DBRX via NVIDIA NIM
-  'databricks/dbrx-instruct': 131_072,
-  // Jamba via NVIDIA NIM
-  'ai21labs/jamba-1.5-large-instruct': 256_000,
-  'ai21labs/jamba-1.5-mini-instruct': 256_000,
-  // Yi via NVIDIA NIM
-  '01-ai/yi-large': 32_768,
-
  // MiniMax (all M2.x variants share 204,800 context, 131,072 max output)
  'MiniMax-M2.7':             204_800,
  'MiniMax-M2.7-highspeed':   204_800,
@@ -169,23 +118,14 @@ const OPENAI_CONTEXT_WINDOWS: Record<string, number> = {
  'minimax-m2.1':             204_800,
  'minimax-m2.1-highspeed':   204_800,

-  // MiniMax new models
-  'MiniMax-Text-01':          524_288,
-  'MiniMax-Text-01-Preview':  262_144,
-  'MiniMax-Vision-01':        32_768,
-  'MiniMax-Vision-01-Fast':   16_384,
-  'MiniMax-M2':               204_800,
-
  // Google (via OpenRouter)
  'google/gemini-2.0-flash':1_048_576,
  'google/gemini-2.5-pro':  1_048_576,

  // Google (native via CLAUDE_CODE_USE_GEMINI)
-  'gemini-2.0-flash':              1_048_576,
-  'gemini-2.5-pro':                1_048_576,
-  'gemini-2.5-flash':              1_048_576,
-  'gemini-3.1-pro':                1_048_576,
-  'gemini-3.1-flash-lite-preview': 1_048_576,
+  'gemini-2.0-flash':       1_048_576,
+  'gemini-2.5-pro':         1_048_576,
+  'gemini-2.5-flash':       1_048_576,

  // Ollama local models
  // Llama 3.1+ models support 128k context natively (Meta official specs).
@@ -204,21 +144,6 @@ const OPENAI_CONTEXT_WINDOWS: Record<string, number> = {
  'llama3.2:1b':              128_000,
  'qwen3:8b':                 128_000,
  'codestral':                 32_768,
-
-  // Alibaba DashScope (Coding Plan)
-  // Model context windows from DashScope API /models endpoint (April 2026).
-  // Values sourced from: qwen3.5-plus/qwen3-coder-plus (1M), qwen3-coder-next/max (256K),
-  // kimi-k2.5 (256K), glm-5/glm-4.7 (198K).
-  // Max output tokens: Qwen variants (64K/32K), GLM (16K).
-  'qwen3.6-plus':           1_000_000,
-  'qwen3.5-plus':           1_000_000,
-  'qwen3-coder-plus':       1_000_000,
-  'qwen3-coder-next':         262_144,
-  'qwen3-max':                262_144,
-  'qwen3-max-2026-01-23':     262_144,
-  'kimi-k2.5':                262_144,
-  'glm-5':                    202_752,
-  'glm-4.7':                  202_752,
 }

 /**
@@ -321,23 +246,15 @@ const OPENAI_MAX_OUTPUT_TOKENS: Record<string, number> = {
  'minimax-m2.5-highspeed':  131_072,
  'minimax-m2.1':            131_072,
  'minimax-m2.1-highspeed':  131_072,
-  // New MiniMax models
-  'MiniMax-M2':              131_072,
-  'MiniMax-Text-01':          65_536,
-  'MiniMax-Text-01-Preview':  65_536,
-  'MiniMax-Vision-01':        16_384,
-  'MiniMax-Vision-01-Fast':    16_384,

  // Google (via OpenRouter)
  'google/gemini-2.0-flash':   8_192,
  'google/gemini-2.5-pro':    65_536,

  // Google (native via CLAUDE_CODE_USE_GEMINI)
-  'gemini-2.0-flash':              8_192,
-  'gemini-2.5-pro':                65_536,
-  'gemini-2.5-flash':              65_536,
-  'gemini-3.1-pro':                65_536,
-  'gemini-3.1-flash-lite-preview': 65_536,
+  'gemini-2.0-flash':          8_192,
+  'gemini-2.5-pro':           65_536,
+  'gemini-2.5-flash':         65_536,

  // Ollama local models (conservative safe defaults)
  'llama3.3:70b':               4_096,
@@ -354,43 +271,6 @@ const OPENAI_MAX_OUTPUT_TOKENS: Record<string, number> = {
  'llama3.2:1b':                4_096,
  'qwen3:8b':                   8_192,
  'codestral':                   8_192,
-
-  // NVIDIA NIM models
-  'nvidia/llama-3.1-nemotron-70b-instruct': 32_768,
-  'nvidia/nemotron-mini-4b-instruct': 8_192,
-  'meta/llama-3.1-405b-instruct': 32_768,
-  'meta/llama-3.1-70b-instruct': 32_768,
-  'meta/llama-3.2-90b-instruct': 32_768,
-  'meta/llama-3.3-70b-instruct': 32_768,
-  'google/gemma-2-27b-it': 4_096,
-  'google/gemma-3-27b-it': 16_384,
-  'google/gemma-3-12b-it': 16_384,
-  'deepseek-ai/deepseek-r1': 32_768,
-  'deepseek-ai/deepseek-v3': 32_768,
-  'deepseek-ai/deepseek-v3.2': 32_768,
-  'qwen/qwen3-32b': 32_768,
-  'qwen/qwen2.5-7b-instruct': 8_192,
-  'mistralai/mistral-large-3-675b-instruct-2512': 32_768,
-  'mistralai/mixtral-8x7b-instruct-v0.1': 8_192,
-  'microsoft/phi-4-mini-instruct': 4_096,
-  'microsoft/phi-3.5-mini-instruct': 4_096,
-  'ibm/granite-3.3-8b-instruct': 4_096,
-  'z-ai/glm5': 32_768,
-  'moonshotai/kimi-k2.5': 32_768,
-  'databricks/dbrx-instruct': 32_768,
-  'ai21labs/jamba-1.5-large-instruct': 32_768,
-  '01-ai/yi-large': 8_192,
-
-  // Alibaba DashScope (Coding Plan)
-  'qwen3.6-plus':              65_536,
-  'qwen3.5-plus':              65_536,
-  'qwen3-coder-plus':          65_536,
-  'qwen3-coder-next':          65_536,
-  'qwen3-max':                 32_768,
-  'qwen3-max-2026-01-23':      32_768,
-  'kimi-k2.5':                 32_768,
-  'glm-5':                     16_384,
-  'glm-4.7':                   16_384,
 }

 function lookupByModel<T>(table: Record<string, T>, model: string): T | undefined {
--- a/src/utils/model/providers.ts
+++ b/src/utils/model/providers.ts
@@ -11,17 +11,9 @@ export type APIProvider =
  | 'gemini'
  | 'github'
  | 'codex'
-  | 'nvidia-nim'
-  | 'minimax'
  | 'mistral'

 export function getAPIProvider(): APIProvider {
-  if (isEnvTruthy(process.env.NVIDIA_NIM)) {
-    return 'nvidia-nim'
-  }
-  if (isEnvTruthy(process.env.MINIMAX_API_KEY)) {
-    return 'minimax'
-  }
  return isEnvTruthy(process.env.CLAUDE_CODE_USE_GEMINI)
    ? 'gemini'
    :
--- a/src/utils/model/validateModel.ts
+++ b/src/utils/model/validateModel.ts
@@ -11,8 +11,6 @@ import {
 } from '@anthropic-ai/sdk'
 import { getModelStrings } from './modelStrings.js'
 import { getCachedOllamaModelOptions, isOllamaProvider } from './ollamaModels.js'
-import { getCachedNvidiaNimModelOptions, isNvidiaNimProvider } from './nvidiaNimModels.js'
-import { getCachedMiniMaxModelOptions, isMiniMaxProvider } from './minimaxModels.js'

 // Cache valid models to avoid repeated API calls
 const validModelCache = new Map<string, boolean>()
@@ -49,40 +47,6 @@ export async function validateModel(
    // If cache is empty, fall through to API validation
  }

-  // For NVIDIA NIM provider, validate against cached model list
-  if (isNvidiaNimProvider()) {
-    const nvidiaModels = getCachedNvidiaNimModelOptions()
-    const found = nvidiaModels.some(m => m.value === normalizedModel)
-    if (found) {
-      validModelCache.set(normalizedModel, true)
-      return { valid: true }
-    }
-    if (nvidiaModels.length > 0) {
-      const MAX_SHOWN = 5
-      const names = nvidiaModels.map(m => m.value)
-      const shown = names.slice(0, MAX_SHOWN).join(', ')
-      const suffix = names.length > MAX_SHOWN ? ` and ${names.length - MAX_SHOWN} more` : ''
-      return { valid: false, error: `Model '${normalizedModel}' not found in NVIDIA NIM catalog. Available: ${shown}${suffix}` }
-    }
-  }
-
-  // For MiniMax provider, validate against cached model list
-  if (isMiniMaxProvider()) {
-    const minimaxModels = getCachedMiniMaxModelOptions()
-    const found = minimaxModels.some(m => m.value === normalizedModel)
-    if (found) {
-      validModelCache.set(normalizedModel, true)
-      return { valid: true }
-    }
-    if (minimaxModels.length > 0) {
-      const MAX_SHOWN = 5
-      const names = minimaxModels.map(m => m.value)
-      const shown = names.slice(0, MAX_SHOWN).join(', ')
-      const suffix = names.length > MAX_SHOWN ? ` and ${names.length - MAX_SHOWN} more` : ''
-      return { valid: false, error: `Model '${normalizedModel}' not found in MiniMax catalog. Available: ${shown}${suffix}` }
-    }
-  }
-
  // Check against availableModels allowlist before any API call
  if (!isModelAllowed(normalizedModel)) {
    return {
--- a/src/utils/openclaudePaths.test.ts
+++ b/src/utils/openclaudePaths.test.ts
@@ -76,9 +76,7 @@ describe('OpenClaude paths', () => {
  })

  test('local installer uses openclaude wrapper path', async () => {
-    // Force .openclaude config home so the test doesn't fall back to
-    // ~/.claude when ~/.openclaude doesn't exist on this machine.
-    process.env.CLAUDE_CONFIG_DIR = join(homedir(), '.openclaude')
+    delete process.env.CLAUDE_CONFIG_DIR
    const { getLocalClaudePath } = await importFreshLocalInstaller()

    expect(getLocalClaudePath()).toBe(
--- a/src/utils/providerDiscovery.ts
+++ b/src/utils/providerDiscovery.ts
@@ -105,14 +105,6 @@ export function getLocalOpenAICompatibleProviderLabel(baseUrl?: string): string
    ) {
      return 'text-generation-webui'
    }
-    // Check for NVIDIA NIM
-    if (host.includes('nvidia') || haystack.includes('nvidia') || host.includes('integrate.api.nvidia')) {
-      return 'NVIDIA NIM'
-    }
-    // Check for MiniMax (both api.minimax.io and api.minimax.chat)
-    if (host.includes('minimax') || haystack.includes('minimax')) {
-      return 'MiniMax'
-    }
  } catch {
    // Fall back to the generic label when the base URL is malformed.
  }
--- a/src/utils/providerFlag.ts
+++ b/src/utils/providerFlag.ts
@@ -21,8 +21,6 @@ export const VALID_PROVIDERS = [
  'bedrock',
  'vertex',
  'ollama',
-  'nvidia-nim',
-  'minimax',
 ] as const

 export type ProviderFlagName = (typeof VALID_PROVIDERS)[number]
@@ -133,21 +131,6 @@ export function applyProviderFlag(
      }
      if (model) process.env.OPENAI_MODEL = model
      break
-
-    case 'nvidia-nim':
-      process.env.CLAUDE_CODE_USE_OPENAI = '1'
-      process.env.OPENAI_BASE_URL ??= 'https://integrate.api.nvidia.com/v1'
-      process.env.NVIDIA_NIM = '1'
-      process.env.OPENAI_MODEL ??= 'nvidia/llama-3.1-nemotron-70b-instruct'
-      if (model) process.env.OPENAI_MODEL = model
-      break
-
-    case 'minimax':
-      process.env.CLAUDE_CODE_USE_OPENAI = '1'
-      process.env.OPENAI_BASE_URL ??= 'https://api.minimax.io/v1'
-      process.env.OPENAI_MODEL ??= 'MiniMax-M2.5'
-      if (model) process.env.OPENAI_MODEL = model
-      break
  }

  return {}
--- a/src/utils/providerModels.test.ts
+++ b/src/utils/providerModels.test.ts
@@ -1,108 +0,0 @@
-import { describe, expect, test } from 'bun:test'
-
-import {
-  getPrimaryModel,
-  hasMultipleModels,
-  parseModelList,
-} from './providerModels.ts'
-
-// ── parseModelList ────────────────────────────────────────────────────────────
-
-describe('parseModelList', () => {
-  test('splits comma-separated models', () => {
-    expect(parseModelList('glm-4.7, glm-4.7-flash')).toEqual([
-      'glm-4.7',
-      'glm-4.7-flash',
-    ])
-  })
-
-  test('returns single model in an array', () => {
-    expect(parseModelList('llama3.1:8b')).toEqual(['llama3.1:8b'])
-  })
-
-  test('trims whitespace around each model', () => {
-    expect(parseModelList('  gpt-4o ,  gpt-4o-mini  , o3-mini ')).toEqual([
-      'gpt-4o',
-      'gpt-4o-mini',
-      'o3-mini',
-    ])
-  })
-
-  test('filters out empty entries from trailing commas', () => {
-    expect(parseModelList('gpt-4o,,gpt-4o-mini,')).toEqual([
-      'gpt-4o',
-      'gpt-4o-mini',
-    ])
-  })
-
-  test('returns empty array for empty string', () => {
-    expect(parseModelList('')).toEqual([])
-  })
-
-  test('returns empty array for whitespace-only string', () => {
-    expect(parseModelList('   ')).toEqual([])
-  })
-
-  test('returns empty array for comma-only string', () => {
-    expect(parseModelList(',,,')).toEqual([])
-  })
-
-  test('handles models with colons', () => {
-    expect(parseModelList('qwen2.5-coder:7b, llama3.1:8b')).toEqual([
-      'qwen2.5-coder:7b',
-      'llama3.1:8b',
-    ])
-  })
-})
-
-// ── getPrimaryModel ───────────────────────────────────────────────────────────
-
-describe('getPrimaryModel', () => {
-  test('returns first model from comma-separated list', () => {
-    expect(getPrimaryModel('glm-4.7, glm-4.7-flash')).toBe('glm-4.7')
-  })
-
-  test('returns the only model when single model is provided', () => {
-    expect(getPrimaryModel('llama3.1:8b')).toBe('llama3.1:8b')
-  })
-
-  test('returns the original string when input is empty', () => {
-    expect(getPrimaryModel('')).toBe('')
-  })
-
-  test('returns first model after trimming', () => {
-    expect(getPrimaryModel('  gpt-4o , gpt-4o-mini')).toBe('gpt-4o')
-  })
-
-  test('returns first model when others are empty from trailing commas', () => {
-    expect(getPrimaryModel('claude-sonnet-4-6,,')).toBe('claude-sonnet-4-6')
-  })
-})
-
-// ── hasMultipleModels ─────────────────────────────────────────────────────────
-
-describe('hasMultipleModels', () => {
-  test('returns true when multiple models are present', () => {
-    expect(hasMultipleModels('glm-4.7, glm-4.7-flash')).toBe(true)
-  })
-
-  test('returns false for a single model', () => {
-    expect(hasMultipleModels('llama3.1:8b')).toBe(false)
-  })
-
-  test('returns false for empty string', () => {
-    expect(hasMultipleModels('')).toBe(false)
-  })
-
-  test('returns false for whitespace-only string', () => {
-    expect(hasMultipleModels('   ')).toBe(false)
-  })
-
-  test('returns false when extra commas produce no extra models', () => {
-    expect(hasMultipleModels('gpt-4o,,')).toBe(false)
-  })
-
-  test('returns true for three models', () => {
-    expect(hasMultipleModels('a, b, c')).toBe(true)
-  })
-})
--- a/src/utils/providerModels.ts
+++ b/src/utils/providerModels.ts
@@ -1,33 +0,0 @@
-/**
- * Utility functions for parsing comma-separated model names in provider profiles.
- *
- * Example: "glm-4.7, glm-4.7-flash" -> ["glm-4.7", "glm-4.7-flash"]
- * Single model: "llama3.1:8b" -> ["llama3.1:8b"]
- */
-
-/**
- * Splits a comma-separated model field into an array of trimmed model names,
- * filtering out any empty entries.
- */
-export function parseModelList(modelField: string): string[] {
-  return modelField
-    .split(',')
-    .map((part) => part.trim())
-    .filter((part) => part.length > 0)
-}
-
-/**
- * Returns the first (primary) model from a comma-separated model field.
- * Falls back to the original string if parsing yields no results.
- */
-export function getPrimaryModel(modelField: string): string {
-  const models = parseModelList(modelField)
-  return models.length > 0 ? models[0] : modelField
-}
-
-/**
- * Returns true if the model field contains more than one model.
- */
-export function hasMultipleModels(modelField: string): boolean {
-  return parseModelList(modelField).length > 1
-}
--- a/src/utils/providerProfile.test.ts
+++ b/src/utils/providerProfile.test.ts
@@ -166,7 +166,7 @@ test('matching persisted gemini env is reused for gemini launch', async () => {
  assert.equal(env.GEMINI_BASE_URL, 'https://example.test/v1beta/openai')
 })

-test('openai env variables take precedence over gemini', async () => {
+test('gemini launch ignores mismatched persisted openai env and strips other provider secrets', async () => {
  const env = await buildLaunchEnv({
    profile: 'gemini',
    persisted: profile('openai', {
@@ -187,16 +187,16 @@ test('openai env variables take precedence over gemini', async () => {
    },
  })

-  assert.equal(env.CLAUDE_CODE_USE_GEMINI, undefined) 
-  assert.equal(env.CLAUDE_CODE_USE_OPENAI, '1')
-  assert.equal(env.GEMINI_MODEL, undefined)
-  assert.equal(env.GEMINI_API_KEY, undefined)
+  assert.equal(env.CLAUDE_CODE_USE_GEMINI, '1')
+  assert.equal(env.CLAUDE_CODE_USE_OPENAI, undefined)
+  assert.equal(env.GEMINI_MODEL, 'gemini-2.0-flash')
+  assert.equal(env.GEMINI_API_KEY, 'gem-live')
  assert.equal(
    env.GEMINI_BASE_URL,
-    undefined,
+    'https://generativelanguage.googleapis.com/v1beta/openai',
  )
  assert.equal(env.GOOGLE_API_KEY, undefined)
-  assert.equal(env.OPENAI_API_KEY, 'sk-live')
+  assert.equal(env.OPENAI_API_KEY, undefined)
  assert.equal(env.CODEX_API_KEY, undefined)
  assert.equal(env.CHATGPT_ACCOUNT_ID, undefined)
 })
@@ -562,13 +562,8 @@ test('buildStartupEnvFromProfile leaves explicit provider selections untouched',
    processEnv,
  })

-  // Remove the strict object equality check: assert.equal(env, processEnv)
+  assert.equal(env, processEnv)
  assert.equal(env.CLAUDE_CODE_USE_GEMINI, '1')
-  assert.equal(env.GEMINI_API_KEY, 'gem-live')
-  assert.equal(env.GEMINI_MODEL, 'gemini-2.0-flash')
-  // Add the new default fields injected by the function
-  assert.equal(env.GEMINI_BASE_URL, 'https://generativelanguage.googleapis.com/v1beta/openai')
-  assert.equal(env.GEMINI_AUTH_MODE, 'api-key')
  assert.equal(env.OPENAI_API_KEY, undefined)
 })

@@ -612,17 +607,14 @@ test('buildStartupEnvFromProfile treats explicit falsey provider flags as user i
    processEnv,
  })

-  assert.equal(env.CLAUDE_CODE_USE_OPENAI, undefined)
-  assert.equal(env.CLAUDE_CODE_USE_GEMINI, '1')
-  assert.equal(env.GEMINI_API_KEY, 'gem-persisted')
-  assert.equal(env.GEMINI_MODEL, 'gemini-2.5-flash')
-  assert.equal(env.GEMINI_BASE_URL, 'https://generativelanguage.googleapis.com/v1beta/openai')
-  assert.equal(env.GEMINI_AUTH_MODE, 'api-key')
+  assert.equal(env, processEnv)
+  assert.equal(env.CLAUDE_CODE_USE_OPENAI, '0')
+  assert.equal(env.GEMINI_API_KEY, undefined)
 })

 test('maskSecretForDisplay preserves only a short prefix and suffix', () => {
-  assert.equal(maskSecretForDisplay('sk-secret-12345678'), 'sk-...678')
-  assert.equal(maskSecretForDisplay('AIzaSecret12345678'), 'AIz...678')
+  assert.equal(maskSecretForDisplay('sk-secret-12345678'), 'sk-...5678')
+  assert.equal(maskSecretForDisplay('AIzaSecret12345678'), 'AIza...5678')
 })

 test('redactSecretValueForDisplay masks poisoned display fields that equal configured secrets', () => {
@@ -630,7 +622,7 @@ test('redactSecretValueForDisplay masks poisoned display fields that equal confi

  assert.equal(
    redactSecretValueForDisplay(apiKey, { OPENAI_API_KEY: apiKey }),
-    'sk-...678',
+    'sk-...5678',
  )
  assert.equal(
    redactSecretValueForDisplay('gpt-4o', { OPENAI_API_KEY: apiKey }),
--- a/src/utils/providerProfile.ts
+++ b/src/utils/providerProfile.ts
@@ -6,32 +6,29 @@ import {
  isCodexBaseUrl,
  resolveCodexApiCredentials,
  resolveProviderRequest,
-} from '../services/api/providerConfig.js'
+} from '../services/api/providerConfig.ts'
 import { parseChatgptAccountId } from '../services/api/codexOAuthShared.js'
 import {
  getGoalDefaultOpenAIModel,
  normalizeRecommendationGoal,
  type RecommendationGoal,
-} from './providerRecommendation.js'
-import { readGeminiAccessToken } from './geminiCredentials.js'
-import { getOllamaChatBaseUrl } from './providerDiscovery.js'
-import { getProviderValidationError } from './providerValidation.js'
+} from './providerRecommendation.ts'
+import { readGeminiAccessToken } from './geminiCredentials.ts'
+import { getOllamaChatBaseUrl } from './providerDiscovery.ts'
+import { getProviderValidationError } from './providerValidation.ts'
 import {
  maskSecretForDisplay,
  redactSecretValueForDisplay,
  sanitizeApiKey,
  sanitizeProviderConfigValue,
-} from './providerSecrets.js'
+} from './providerSecrets.ts'

 export {
  maskSecretForDisplay,
  redactSecretValueForDisplay,
  sanitizeApiKey,
  sanitizeProviderConfigValue,
-} from './providerSecrets.js'
-import { isEnvTruthy } from './envUtils.ts'
-
-import { PROVIDERS } from './configConstants.js'
+} from './providerSecrets.ts'

 export const PROFILE_FILE_NAME = '.openclaude-profile.json'
 export const DEFAULT_GEMINI_BASE_URL =
@@ -60,28 +57,18 @@ const PROFILE_ENV_KEYS = [
  'GEMINI_MODEL',
  'GEMINI_BASE_URL',
  'GOOGLE_API_KEY',
-  'NVIDIA_NIM',
-  'NVIDIA_API_KEY',
-  'NVIDIA_MODEL',
-  'MINIMAX_API_KEY',
-  'MINIMAX_BASE_URL',
-  'MINIMAX_MODEL',
  'MISTRAL_BASE_URL',
  'MISTRAL_API_KEY',
  'MISTRAL_MODEL',
 ] as const

-const SECRET_ENV_KEYS = [
-  'OPENAI_API_KEY',
-  'CODEX_API_KEY',
-  'GEMINI_API_KEY',
-  'GOOGLE_API_KEY',
-  'NVIDIA_API_KEY',
-  'MINIMAX_API_KEY',
-  'MISTRAL_API_KEY',
-] as const
-
-export type ProviderProfile = 'openai' | 'ollama' | 'codex' | 'gemini' | 'atomic-chat' | 'nvidia-nim' | 'minimax' | 'mistral'
+export type ProviderProfile =
+  | 'openai'
+  | 'ollama'
+  | 'codex'
+  | 'gemini'
+  | 'atomic-chat'
+  | 'mistral'

 export type ProfileEnv = {
  OPENAI_BASE_URL?: string
@@ -95,12 +82,6 @@ export type ProfileEnv = {
  GEMINI_AUTH_MODE?: 'api-key' | 'access-token' | 'adc'
  GEMINI_MODEL?: string
  GEMINI_BASE_URL?: string
-  GOOGLE_API_KEY?: string
-  NVIDIA_NIM?: string
-  NVIDIA_API_KEY?: string
-  MINIMAX_API_KEY?: string
-  MINIMAX_BASE_URL?: string
-  MINIMAX_MODEL?: string
  MISTRAL_BASE_URL?: string
  MISTRAL_API_KEY?: string
  MISTRAL_MODEL?: string
@@ -112,19 +93,6 @@ export type ProfileFile = {
  createdAt: string
 }

-type SecretValueSource = Partial<
-  Record<
-    | 'OPENAI_API_KEY'
-    | 'CODEX_API_KEY'
-    | 'GEMINI_API_KEY'
-    | 'GOOGLE_API_KEY'
-    | 'NVIDIA_API_KEY'
-    | 'MINIMAX_API_KEY'
-    | 'MISTRAL_API_KEY',
-    string | undefined
-  >
->
-
 type ProfileFileLocation = {
  cwd?: string
  filePath?: string
@@ -145,8 +113,6 @@ export function isProviderProfile(value: unknown): value is ProviderProfile {
    value === 'codex' ||
    value === 'gemini' ||
    value === 'atomic-chat' ||
-    value === 'nvidia-nim' ||
-    value === 'minimax' ||
    value === 'mistral'
  )
 }
@@ -177,67 +143,6 @@ export function buildAtomicChatProfileEnv(
  }
 }

-export function buildNvidiaNimProfileEnv(options: {
-  model?: string | null
-  baseUrl?: string | null
-  apiKey?: string | null
-  processEnv?: NodeJS.ProcessEnv
-}): ProfileEnv | null {
-  const processEnv = options.processEnv ?? process.env
-  const key = sanitizeApiKey(options.apiKey ?? processEnv.NVIDIA_API_KEY)
-  if (!key) {
-    return null
-  }
-
-  const defaultBaseUrl = 'https://integrate.api.nvidia.com/v1'
-  const secretSource: SecretValueSource = { OPENAI_API_KEY: key }
-
-  return {
-    OPENAI_BASE_URL:
-      sanitizeProviderConfigValue(options.baseUrl, secretSource) ||
-      sanitizeProviderConfigValue(processEnv.OPENAI_BASE_URL, secretSource) ||
-      defaultBaseUrl,
-    OPENAI_MODEL:
-      sanitizeProviderConfigValue(options.model, secretSource) ||
-      sanitizeProviderConfigValue(processEnv.OPENAI_MODEL, secretSource) ||
-      'nvidia/llama-3.1-nemotron-70b-instruct',
-    OPENAI_API_KEY: key,
-    NVIDIA_NIM: '1',
-  }
-}
-
-export function buildMiniMaxProfileEnv(options: {
-  model?: string | null
-  baseUrl?: string | null
-  apiKey?: string | null
-  processEnv?: NodeJS.ProcessEnv
-}): ProfileEnv | null {
-  const processEnv = options.processEnv ?? process.env
-  const key = sanitizeApiKey(options.apiKey ?? processEnv.MINIMAX_API_KEY)
-  if (!key) {
-    return null
-  }
-
-  const defaultBaseUrl = 'https://api.minimax.io/v1'
-  const defaultModel = 'MiniMax-M2.5'
-  const secretSource: SecretValueSource = { OPENAI_API_KEY: key }
-
-  return {
-    OPENAI_BASE_URL:
-      sanitizeProviderConfigValue(options.baseUrl, secretSource) ||
-      sanitizeProviderConfigValue(processEnv.OPENAI_BASE_URL, secretSource) ||
-      defaultBaseUrl,
-    OPENAI_MODEL:
-      sanitizeProviderConfigValue(options.model, secretSource) ||
-      sanitizeProviderConfigValue(processEnv.OPENAI_MODEL, secretSource) ||
-      defaultModel,
-    OPENAI_API_KEY: key,
-    MINIMAX_API_KEY: key,
-    MINIMAX_BASE_URL: defaultBaseUrl,
-    MINIMAX_MODEL: defaultModel,
-  }
-}
-
 export function buildGeminiProfileEnv(options: {
  model?: string | null
  baseUrl?: string | null
@@ -256,13 +161,15 @@ export function buildGeminiProfileEnv(options: {
    return null
  }

-  const secretSource: SecretValueSource = key ? { GEMINI_API_KEY: key } : {}
-
  const env: ProfileEnv = {
    GEMINI_AUTH_MODE: authMode,
    GEMINI_MODEL:
-      sanitizeProviderConfigValue(options.model, secretSource) ||
-      sanitizeProviderConfigValue(processEnv.GEMINI_MODEL, secretSource) ||
+      sanitizeProviderConfigValue(options.model, { GEMINI_API_KEY: key }, processEnv) ||
+      sanitizeProviderConfigValue(
+        processEnv.GEMINI_MODEL,
+        { GEMINI_API_KEY: key },
+        processEnv,
+      ) ||
      DEFAULT_GEMINI_MODEL,
  }

@@ -271,8 +178,12 @@ export function buildGeminiProfileEnv(options: {
  }

  const baseUrl =
-    sanitizeProviderConfigValue(options.baseUrl, secretSource) ||
-    sanitizeProviderConfigValue(processEnv.GEMINI_BASE_URL, secretSource)
+    sanitizeProviderConfigValue(options.baseUrl, { GEMINI_API_KEY: key }, processEnv) ||
+    sanitizeProviderConfigValue(
+      processEnv.GEMINI_BASE_URL,
+      { GEMINI_API_KEY: key },
+      processEnv,
+    )
  if (baseUrl) {
    env.GEMINI_BASE_URL = baseUrl
  }
@@ -294,14 +205,15 @@ export function buildOpenAIProfileEnv(options: {
  }

  const defaultModel = getGoalDefaultOpenAIModel(options.goal)
-  const secretSource: SecretValueSource = { OPENAI_API_KEY: key }
  const shellOpenAIModel = sanitizeProviderConfigValue(
    processEnv.OPENAI_MODEL,
-    secretSource,
+    { OPENAI_API_KEY: key },
+    processEnv,
  )
  const shellOpenAIBaseUrl = sanitizeProviderConfigValue(
    processEnv.OPENAI_BASE_URL,
-    secretSource,
+    { OPENAI_API_KEY: key },
+    processEnv,
  )
  const shellOpenAIRequest = resolveProviderRequest({
    model: shellOpenAIModel,
@@ -312,11 +224,19 @@ export function buildOpenAIProfileEnv(options: {

  return {
    OPENAI_BASE_URL:
-      sanitizeProviderConfigValue(options.baseUrl, secretSource) ||
+      sanitizeProviderConfigValue(
+        options.baseUrl,
+        { OPENAI_API_KEY: key },
+        processEnv,
+      ) ||
      (useShellOpenAIConfig ? shellOpenAIBaseUrl : undefined) ||
      DEFAULT_OPENAI_BASE_URL,
    OPENAI_MODEL:
-      sanitizeProviderConfigValue(options.model, secretSource) ||
+      sanitizeProviderConfigValue(
+        options.model,
+        { OPENAI_API_KEY: key },
+        processEnv,
+      ) ||
      (useShellOpenAIConfig ? shellOpenAIModel : undefined) ||
      defaultModel,
    OPENAI_API_KEY: key,
@@ -373,19 +293,21 @@ export function buildMistralProfileEnv(options: {
  const env: ProfileEnv = {
    MISTRAL_API_KEY: key,
    MISTRAL_MODEL:
-      sanitizeProviderConfigValue(options.model, { MISTRAL_API_KEY: key }) ||
+      sanitizeProviderConfigValue(options.model, { MISTRAL_API_KEY: key }, processEnv) ||
      sanitizeProviderConfigValue(
        processEnv.MISTRAL_MODEL,
        { MISTRAL_API_KEY: key },
+        processEnv,
      ) ||
      DEFAULT_MISTRAL_MODEL,
  }

  const baseUrl =
-    sanitizeProviderConfigValue(options.baseUrl, { MISTRAL_API_KEY: key }) ||
+    sanitizeProviderConfigValue(options.baseUrl, { MISTRAL_API_KEY: key }, processEnv) ||
    sanitizeProviderConfigValue(
      processEnv.MISTRAL_BASE_URL,
      { MISTRAL_API_KEY: key },
+      processEnv,
    )
  if (baseUrl) {
    env.MISTRAL_BASE_URL = baseUrl
@@ -501,13 +423,13 @@ export function hasExplicitProviderSelection(
  }

  return (
-    isEnvTruthy(processEnv.CLAUDE_CODE_USE_OPENAI) ||
-    isEnvTruthy(processEnv.CLAUDE_CODE_USE_GITHUB) ||
-    isEnvTruthy(processEnv.CLAUDE_CODE_USE_GEMINI) ||
-    isEnvTruthy(processEnv.CLAUDE_CODE_USE_MISTRAL) ||
-    isEnvTruthy(processEnv.CLAUDE_CODE_USE_BEDROCK) ||
-    isEnvTruthy(processEnv.CLAUDE_CODE_USE_VERTEX) ||
-    isEnvTruthy(processEnv.CLAUDE_CODE_USE_FOUNDRY)
+    processEnv.CLAUDE_CODE_USE_OPENAI !== undefined ||
+    processEnv.CLAUDE_CODE_USE_GITHUB !== undefined ||
+    processEnv.CLAUDE_CODE_USE_GEMINI !== undefined ||
+    processEnv.CLAUDE_CODE_USE_MISTRAL !== undefined ||
+    processEnv.CLAUDE_CODE_USE_BEDROCK !== undefined ||
+    processEnv.CLAUDE_CODE_USE_VERTEX !== undefined ||
+    processEnv.CLAUDE_CODE_USE_FOUNDRY !== undefined
  )
 }

@@ -543,11 +465,11 @@ export async function buildLaunchEnv(options: {
  )
  const shellOpenAIModel = sanitizeProviderConfigValue(
    processEnv.OPENAI_MODEL,
-    processEnv as SecretValueSource,
+    processEnv,
  )
  const shellOpenAIBaseUrl = sanitizeProviderConfigValue(
    processEnv.OPENAI_BASE_URL,
-    processEnv as SecretValueSource,
+    processEnv,
  )
  const persistedGeminiModel = sanitizeProviderConfigValue(
    persistedEnv.GEMINI_MODEL,
@@ -559,11 +481,11 @@ export async function buildLaunchEnv(options: {
  )
  const shellGeminiModel = sanitizeProviderConfigValue(
    processEnv.GEMINI_MODEL,
-    processEnv as SecretValueSource,
+    processEnv,
  )
  const shellGeminiBaseUrl = sanitizeProviderConfigValue(
    processEnv.GEMINI_BASE_URL,
-    processEnv as SecretValueSource,
+    processEnv,
  )
  const shellGeminiAccessToken =
    processEnv.GEMINI_ACCESS_TOKEN?.trim() || undefined
@@ -576,20 +498,6 @@ export async function buildLaunchEnv(options: {
  const persistedGeminiKey = sanitizeApiKey(persistedEnv.GEMINI_API_KEY)
  const persistedGeminiAuthMode = persistedEnv.GEMINI_AUTH_MODE

-  if (hasExplicitProviderSelection(processEnv)) {
-    for (let provider of PROVIDERS) {
-      if (provider === "anthropic") {
-        continue;
-      }
-
-      const env_key_name = `CLAUDE_CODE_USE_${provider.toUpperCase()}`
-
-      if (env_key_name in processEnv && isEnvTruthy(processEnv[env_key_name])) {
-        options.profile = provider;
-      }
-    }
-  }
-
  if (options.profile === 'gemini') {
    const env: NodeJS.ProcessEnv = {
      ...processEnv,
@@ -659,15 +567,19 @@ export async function buildLaunchEnv(options: {

    const shellMistralModel = sanitizeProviderConfigValue(
      processEnv.MISTRAL_MODEL,
+      processEnv,
    )
    const persistedMistralModel = sanitizeProviderConfigValue(
      persistedEnv.MISTRAL_MODEL,
+      persistedEnv,
    )
    const shellMistralBaseUrl = sanitizeProviderConfigValue(
      processEnv.MISTRAL_BASE_URL,
+      processEnv,
    )
    const persistedMistralBaseUrl = sanitizeProviderConfigValue(
      persistedEnv.MISTRAL_BASE_URL,
+      persistedEnv,
    )

    env.MISTRAL_MODEL =
@@ -842,18 +754,12 @@ export async function buildStartupEnvFromProfile(options?: {
  const persisted = options?.persisted ?? loadProfileFile()

  // Saved /provider profiles should still win over provider-manager env that was
-  // auto-applied during startup. Only an explicit shell/flag provider selection
+  // auto-applied during startup. Only explicit shell/flag provider selection
  // should bypass the persisted startup profile.
-  //
  const profileManagedEnv = processEnv.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED === '1'
-
-  // If the user explicitly selected a provider via env, allow it to bypass
-  // the persisted profile only when we can prove it was managed by the
-  // persisted profile env itself.
-  //
-  // Practically: on initial startup, provider routing env vars can already
-  // be present due to earlier auto-application steps. We should still apply
-  // the persisted profile rather than returning early.
+  if (hasExplicitProviderSelection(processEnv) && !profileManagedEnv) {
+    return processEnv
+  }

  if (!persisted) {
    return processEnv
--- a/src/utils/providerProfiles.test.ts
+++ b/src/utils/providerProfiles.test.ts
@@ -13,7 +13,6 @@ const RESTORED_KEYS = [
  'CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED_ID',
  'CLAUDE_CODE_USE_OPENAI',
  'CLAUDE_CODE_USE_GEMINI',
-  'CLAUDE_CODE_USE_MISTRAL',
  'CLAUDE_CODE_USE_GITHUB',
  'CLAUDE_CODE_USE_BEDROCK',
  'CLAUDE_CODE_USE_VERTEX',
@@ -25,15 +24,6 @@ const RESTORED_KEYS = [
  'ANTHROPIC_BASE_URL',
  'ANTHROPIC_MODEL',
  'ANTHROPIC_API_KEY',
-  'GEMINI_BASE_URL',
-  'GEMINI_MODEL',
-  'GEMINI_API_KEY',
-  'GEMINI_AUTH_MODE',
-  'GEMINI_ACCESS_TOKEN',
-  'GOOGLE_API_KEY',
-  'MISTRAL_BASE_URL',
-  'MISTRAL_MODEL',
-  'MISTRAL_API_KEY',
 ] as const

 type MockConfigState = {
@@ -108,24 +98,6 @@ function buildProfile(overrides: Partial<ProviderProfile> = {}): ProviderProfile
  }
 }

-function buildMistralProfile(overrides: Partial<ProviderProfile> = {}): ProviderProfile {
-  return buildProfile({
-    provider: 'mistral',
-    baseUrl: 'https://api.mistral.ai/v1',
-    model: 'devstral-latest',
-    ...overrides,
-  })
-}
-
-function buildGeminiProfile(overrides: Partial<ProviderProfile> = {}): ProviderProfile {
-  return buildProfile({
-    provider: 'gemini',
-    baseUrl: 'https://generativelanguage.googleapis.com/v1beta/openai',
-    model: 'gemini-3-flash-preview',
-    ...overrides,
-  })
-}
-
 describe('applyProviderProfileToProcessEnv', () => {
  test('openai profile clears competing gemini/github flags', async () => {
    const { applyProviderProfileToProcessEnv } =
@@ -146,36 +118,6 @@ describe('applyProviderProfileToProcessEnv', () => {
    expect(getFreshAPIProvider()).toBe('openai')
  })

-  test('mistral profile sets CLAUDE_CODE_USE_MISTRAL and clears openai flags', async () => {
-    const { applyProviderProfileToProcessEnv } =
-      await importFreshProviderProfileModules()
-    process.env.CLAUDE_CODE_USE_OPENAI = '1'
-
-    applyProviderProfileToProcessEnv(buildMistralProfile())
-    const { getAPIProvider: getFreshAPIProvider } =
-      await importFreshProvidersModule()
-
-    expect(process.env.CLAUDE_CODE_USE_MISTRAL).toBe('1')
-    expect(process.env.CLAUDE_CODE_USE_OPENAI).toBeUndefined()
-    expect(process.env.MISTRAL_MODEL).toBe('devstral-latest')
-    expect(getFreshAPIProvider()).toBe('mistral')
-  })
-
-  test('gemini profile sets CLAUDE_CODE_USE_GEMINI and clears openai flags', async () => {
-    const { applyProviderProfileToProcessEnv } =
-      await importFreshProviderProfileModules()
-    process.env.CLAUDE_CODE_USE_OPENAI = '1'
-
-    applyProviderProfileToProcessEnv(buildGeminiProfile())
-    const { getAPIProvider: getFreshAPIProvider } =
-      await importFreshProvidersModule()
-
-    expect(process.env.CLAUDE_CODE_USE_GEMINI).toBe('1')
-    expect(process.env.CLAUDE_CODE_USE_OPENAI).toBeUndefined()
-    expect(process.env.GEMINI_MODEL).toBe('gemini-3-flash-preview')
-    expect(getFreshAPIProvider()).toBe('gemini')
-  })
-
  test('anthropic profile clears competing gemini/github flags', async () => {
    const { applyProviderProfileToProcessEnv } =
      await importFreshProviderProfileModules()
@@ -197,39 +139,6 @@ describe('applyProviderProfileToProcessEnv', () => {
    expect(process.env.CLAUDE_CODE_USE_OPENAI).toBeUndefined()
    expect(getFreshAPIProvider()).toBe('firstParty')
  })
-
-  test('openai profile with multi-model string sets only first model in OPENAI_MODEL', async () => {
-    const { applyProviderProfileToProcessEnv } =
-      await importFreshProviderProfileModules()
-
-    applyProviderProfileToProcessEnv(
-      buildProfile({
-        provider: 'openai',
-        baseUrl: 'https://api.openai.com/v1',
-        model: 'glm-4.7, glm-4.7-flash, glm-4.7-plus',
-      }),
-    )
-
-    expect(process.env.OPENAI_MODEL).toBe('glm-4.7')
-    expect(String(process.env.CLAUDE_CODE_USE_OPENAI)).toBe('1')
-    expect(process.env.OPENAI_BASE_URL).toBe('https://api.openai.com/v1')
-  })
-
-  test('anthropic profile with multi-model string sets only first model in ANTHROPIC_MODEL', async () => {
-    const { applyProviderProfileToProcessEnv } =
-      await importFreshProviderProfileModules()
-
-    applyProviderProfileToProcessEnv(
-      buildProfile({
-        provider: 'anthropic',
-        baseUrl: 'https://api.anthropic.com',
-        model: 'claude-sonnet-4-6, claude-opus-4-6',
-      }),
-    )
-
-    expect(process.env.ANTHROPIC_MODEL).toBe('claude-sonnet-4-6')
-    expect(process.env.ANTHROPIC_BASE_URL).toBe('https://api.anthropic.com')
-  })
 })

 describe('applyActiveProviderProfileFromConfig', () => {
@@ -452,169 +361,6 @@ describe('getProviderPresetDefaults', () => {
  })
 })

-describe('setActiveProviderProfile', () => {
-  test('sets OPENAI_MODEL env var when switching to an openai-type provider', async () => {
-    const { setActiveProviderProfile } =
-      await importFreshProviderProfileModules()
-    const openaiProfile = buildProfile({
-      id: 'openai_prof',
-      name: 'OpenAI Provider',
-      provider: 'openai',
-      baseUrl: 'https://api.openai.com/v1',
-      model: 'gpt-4o',
-    })
-
-    saveMockGlobalConfig(current => ({
-      ...current,
-      providerProfiles: [openaiProfile],
-    }))
-
-    const result = setActiveProviderProfile('openai_prof')
-
-    expect(result?.id).toBe('openai_prof')
-    expect(String(process.env.CLAUDE_CODE_USE_OPENAI)).toBe('1')
-    expect(process.env.OPENAI_MODEL).toBe('gpt-4o')
-    expect(process.env.OPENAI_BASE_URL).toBe('https://api.openai.com/v1')
-    expect(process.env.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED_ID).toBe(
-      'openai_prof',
-    )
-  })
-
-  test('sets ANTHROPIC_MODEL env var when switching to an anthropic-type provider', async () => {
-    const { setActiveProviderProfile } =
-      await importFreshProviderProfileModules()
-    const anthropicProfile = buildProfile({
-      id: 'anthro_prof',
-      name: 'Anthropic Provider',
-      provider: 'anthropic',
-      baseUrl: 'https://api.anthropic.com',
-      model: 'claude-sonnet-4-6',
-    })
-
-    saveMockGlobalConfig(current => ({
-      ...current,
-      providerProfiles: [anthropicProfile],
-    }))
-
-    const result = setActiveProviderProfile('anthro_prof')
-
-    expect(result?.id).toBe('anthro_prof')
-    expect(process.env.ANTHROPIC_MODEL).toBe('claude-sonnet-4-6')
-    expect(process.env.ANTHROPIC_BASE_URL).toBe('https://api.anthropic.com')
-    expect(process.env.CLAUDE_CODE_USE_OPENAI).toBeUndefined()
-    expect(process.env.OPENAI_MODEL).toBeUndefined()
-    expect(process.env.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED_ID).toBe(
-      'anthro_prof',
-    )
-  })
-
-  test('clears openai model env and sets anthropic model env when switching from openai to anthropic provider', async () => {
-    const { setActiveProviderProfile } =
-      await importFreshProviderProfileModules()
-    const openaiProfile = buildProfile({
-      id: 'openai_prof',
-      name: 'OpenAI Provider',
-      provider: 'openai',
-      baseUrl: 'https://api.openai.com/v1',
-      model: 'gpt-4o',
-      apiKey: 'sk-openai-key',
-    })
-    const anthropicProfile = buildProfile({
-      id: 'anthro_prof',
-      name: 'Anthropic Provider',
-      provider: 'anthropic',
-      baseUrl: 'https://api.anthropic.com',
-      model: 'claude-sonnet-4-6',
-      apiKey: 'sk-ant-key',
-    })
-
-    saveMockGlobalConfig(current => ({
-      ...current,
-      providerProfiles: [openaiProfile, anthropicProfile],
-    }))
-
-    // First activate the openai profile
-    setActiveProviderProfile('openai_prof')
-    expect(process.env.OPENAI_MODEL).toBe('gpt-4o')
-    expect(String(process.env.CLAUDE_CODE_USE_OPENAI)).toBe('1')
-
-    // Now switch to the anthropic profile
-    const result = setActiveProviderProfile('anthro_prof')
-
-    expect(result?.id).toBe('anthro_prof')
-    expect(process.env.ANTHROPIC_MODEL).toBe('claude-sonnet-4-6')
-    expect(process.env.ANTHROPIC_BASE_URL).toBe('https://api.anthropic.com')
-    expect(process.env.CLAUDE_CODE_USE_OPENAI).toBeUndefined()
-    expect(process.env.OPENAI_MODEL).toBeUndefined()
-    expect(process.env.OPENAI_BASE_URL).toBeUndefined()
-    expect(process.env.OPENAI_API_KEY).toBeUndefined()
-    expect(process.env.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED_ID).toBe(
-      'anthro_prof',
-    )
-  })
-
-  test('clears anthropic model env and sets openai model env when switching from anthropic to openai provider', async () => {
-    const { setActiveProviderProfile } =
-      await importFreshProviderProfileModules()
-    const anthropicProfile = buildProfile({
-      id: 'anthro_prof',
-      name: 'Anthropic Provider',
-      provider: 'anthropic',
-      baseUrl: 'https://api.anthropic.com',
-      model: 'claude-sonnet-4-6',
-      apiKey: 'sk-ant-key',
-    })
-    const openaiProfile = buildProfile({
-      id: 'openai_prof',
-      name: 'OpenAI Provider',
-      provider: 'openai',
-      baseUrl: 'https://api.openai.com/v1',
-      model: 'gpt-4o',
-      apiKey: 'sk-openai-key',
-    })
-
-    saveMockGlobalConfig(current => ({
-      ...current,
-      providerProfiles: [anthropicProfile, openaiProfile],
-    }))
-
-    // First activate the anthropic profile
-    setActiveProviderProfile('anthro_prof')
-    expect(process.env.ANTHROPIC_MODEL).toBe('claude-sonnet-4-6')
-    expect(process.env.ANTHROPIC_BASE_URL).toBe('https://api.anthropic.com')
-
-    // Now switch to the openai profile
-    const result = setActiveProviderProfile('openai_prof')
-
-    expect(result?.id).toBe('openai_prof')
-    expect(String(process.env.CLAUDE_CODE_USE_OPENAI)).toBe('1')
-    expect(process.env.OPENAI_MODEL).toBe('gpt-4o')
-    expect(process.env.OPENAI_BASE_URL).toBe('https://api.openai.com/v1')
-    // ANTHROPIC_MODEL is set to the profile model for all provider types
-    expect(process.env.ANTHROPIC_MODEL).toBe('gpt-4o')
-    expect(process.env.ANTHROPIC_BASE_URL).toBeUndefined()
-    expect(process.env.ANTHROPIC_API_KEY).toBeUndefined()
-    expect(process.env.CLAUDE_CODE_PROVIDER_PROFILE_ENV_APPLIED_ID).toBe(
-      'openai_prof',
-    )
-  })
-
-  test('returns null for non-existent profile id', async () => {
-    const { setActiveProviderProfile } =
-      await importFreshProviderProfileModules()
-    const openaiProfile = buildProfile({ id: 'existing_prof' })
-
-    saveMockGlobalConfig(current => ({
-      ...current,
-      providerProfiles: [openaiProfile],
-    }))
-
-    const result = setActiveProviderProfile('nonexistent_prof')
-
-    expect(result).toBeNull()
-  })
-})
-
 describe('deleteProviderProfile', () => {
  test('deleting final profile clears provider env when active profile applied it', async () => {
    const {
@@ -683,82 +429,3 @@ describe('deleteProviderProfile', () => {
    expect(process.env.OPENAI_MODEL).toBe('qwen2.5:3b')
  })
 })
-
-describe('getProfileModelOptions', () => {
-  test('generates options for multi-model profile', async () => {
-    const { getProfileModelOptions } =
-      await importFreshProviderProfileModules()
-
-    const options = getProfileModelOptions(
-      buildProfile({
-        name: 'Test Provider',
-        model: 'glm-4.7, glm-4.7-flash, glm-4.7-plus',
-      }),
-    )
-
-    expect(options).toEqual([
-      { value: 'glm-4.7', label: 'glm-4.7', description: 'Provider: Test Provider' },
-      { value: 'glm-4.7-flash', label: 'glm-4.7-flash', description: 'Provider: Test Provider' },
-      { value: 'glm-4.7-plus', label: 'glm-4.7-plus', description: 'Provider: Test Provider' },
-    ])
-  })
-
-  test('returns single option for single-model profile', async () => {
-    const { getProfileModelOptions } =
-      await importFreshProviderProfileModules()
-
-    const options = getProfileModelOptions(
-      buildProfile({
-        name: 'Single Model',
-        model: 'llama3.1:8b',
-      }),
-    )
-
-    expect(options).toEqual([
-      { value: 'llama3.1:8b', label: 'llama3.1:8b', description: 'Provider: Single Model' },
-    ])
-  })
-
-  test('returns empty array for empty model field', async () => {
-    const { getProfileModelOptions } =
-      await importFreshProviderProfileModules()
-
-    const options = getProfileModelOptions(
-      buildProfile({
-        name: 'Empty',
-        model: '',
-      }),
-    )
-
-    expect(options).toEqual([])
-  })
-})
-
-describe('setActiveProviderProfile model cache', () => {
-  test('populates model cache with all models from multi-model profile on activation', async () => {
-    const {
-      setActiveProviderProfile,
-      getActiveOpenAIModelOptionsCache,
-    } = await importFreshProviderProfileModules()
-
-    mockConfigState = {
-      ...createMockConfigState(),
-      providerProfiles: [
-        buildProfile({
-          id: 'multi_provider',
-          name: 'Multi Provider',
-          model: 'glm-4.7, glm-4.7-flash, glm-4.7-plus',
-          baseUrl: 'https://api.example.com/v1',
-        }),
-      ],
-    }
-
-    setActiveProviderProfile('multi_provider')
-
-    const cache = getActiveOpenAIModelOptionsCache()
-    const cacheValues = cache.map(opt => opt.value)
-    expect(cacheValues).toContain('glm-4.7')
-    expect(cacheValues).toContain('glm-4.7-flash')
-    expect(cacheValues).toContain('glm-4.7-plus')
-  })
-})
--- a/src/utils/providerProfiles.ts
+++ b/src/utils/providerProfiles.ts
@@ -5,15 +5,6 @@ import {
  type ProviderProfile,
 } from './config.js'
 import type { ModelOption } from './model/modelOptions.js'
-import { getPrimaryModel, parseModelList } from './providerModels.js'
-import {
-  createProfileFile,
-  saveProfileFile,
-  buildGeminiProfileEnv,
-  buildMistralProfileEnv,
-  buildOpenAIProfileEnv,
-  type ProviderProfile as ProviderProfileStartup,
-} from './providerProfile.js'

 export type ProviderPreset =
  | 'anthropic'
@@ -28,11 +19,7 @@ export type ProviderPreset =
  | 'azure-openai'
  | 'openrouter'
  | 'lmstudio'
-  | 'dashscope-cn'
-  | 'dashscope-intl'
  | 'custom'
-  | 'nvidia-nim'
-  | 'minimax'

 export type ProviderProfileInput = {
  provider?: ProviderProfile['provider']
@@ -68,14 +55,7 @@ function normalizeBaseUrl(value: string): string {
 function sanitizeProfile(profile: ProviderProfile): ProviderProfile | null {
  const id = trimValue(profile.id)
  const name = trimValue(profile.name)
-  const provider =
-    profile.provider === 'anthropic'
-      ? 'anthropic'
-      : profile.provider === 'mistral'
-        ? 'mistral'
-        : profile.provider === 'gemini'
-          ? 'gemini'
-          : 'openai'
+  const provider = profile.provider === 'anthropic' ? 'anthropic' : 'openai'
  const baseUrl = normalizeBaseUrl(profile.baseUrl)
  const model = trimValue(profile.model)

@@ -176,7 +156,7 @@ export function getProviderPresetDefaults(
      }
    case 'gemini':
      return {
-        provider: 'gemini',
+        provider: 'openai',
        name: 'Google Gemini',
        baseUrl: 'https://generativelanguage.googleapis.com/v1beta/openai',
        model: 'gemini-3-flash-preview',
@@ -185,7 +165,7 @@ export function getProviderPresetDefaults(
      }
    case 'mistral':
      return {
-        provider: 'mistral',
+        provider: 'openai',
        name: 'Mistral',
        baseUrl: 'https://api.mistral.ai/v1',
        model: 'devstral-latest',
@@ -237,24 +217,6 @@ export function getProviderPresetDefaults(
        apiKey: '',
        requiresApiKey: false,
      }
-    case 'dashscope-cn':
-      return {
-        provider: 'openai',
-        name: 'Alibaba Coding Plan (China)',
-        baseUrl: 'https://coding.dashscope.aliyuncs.com/v1',
-        model: 'qwen3.6-plus',
-        apiKey: process.env.DASHSCOPE_API_KEY ?? '',
-        requiresApiKey: true,
-      }
-    case 'dashscope-intl':
-      return {
-        provider: 'openai',
-        name: 'Alibaba Coding Plan',
-        baseUrl: 'https://coding-intl.dashscope.aliyuncs.com/v1',
-        model: 'qwen3.6-plus',
-        apiKey: process.env.DASHSCOPE_API_KEY ?? '',
-        requiresApiKey: true,
-      }
    case 'custom':
      return {
        provider: 'openai',
@@ -267,24 +229,6 @@ export function getProviderPresetDefaults(
        apiKey: process.env.OPENAI_API_KEY ?? '',
        requiresApiKey: false,
      }
-    case 'nvidia-nim':
-      return {
-        provider: 'openai',
-        name: 'NVIDIA NIM',
-        baseUrl: 'https://integrate.api.nvidia.com/v1',
-        model: 'nvidia/llama-3.1-nemotron-70b-instruct',
-        apiKey: process.env.NVIDIA_API_KEY ?? '',
-        requiresApiKey: true,
-      }
-    case 'minimax':
-      return {
-        provider: 'openai',
-        name: 'MiniMax',
-        baseUrl: 'https://api.minimax.io/v1',
-        model: 'MiniMax-M2.5',
-        apiKey: process.env.MINIMAX_API_KEY ?? '',
-        requiresApiKey: true,
-      }
    case 'ollama':
    default:
      return {
@@ -332,7 +276,6 @@ function hasConflictingProviderFlagsForProfile(

  return (
    processEnv.CLAUDE_CODE_USE_GEMINI !== undefined ||
-    processEnv.CLAUDE_CODE_USE_MISTRAL !== undefined ||
    processEnv.CLAUDE_CODE_USE_GITHUB !== undefined ||
    processEnv.CLAUDE_CODE_USE_BEDROCK !== undefined ||
    processEnv.CLAUDE_CODE_USE_VERTEX !== undefined ||
@@ -368,44 +311,12 @@ function isProcessEnvAlignedWithProfile(
    return (
      !hasProviderSelectionFlags(processEnv) &&
      sameOptionalEnvValue(processEnv.ANTHROPIC_BASE_URL, profile.baseUrl) &&
-      sameOptionalEnvValue(processEnv.ANTHROPIC_MODEL, getPrimaryModel(profile.model)) &&
+      sameOptionalEnvValue(processEnv.ANTHROPIC_MODEL, profile.model) &&
      (!includeApiKey ||
        sameOptionalEnvValue(processEnv.ANTHROPIC_API_KEY, profile.apiKey))
    )
  }

-  if (profile.provider === 'mistral') {
-    return (
-      processEnv.CLAUDE_CODE_USE_MISTRAL !== undefined &&
-      processEnv.CLAUDE_CODE_USE_GEMINI === undefined &&
-      processEnv.CLAUDE_CODE_USE_OPENAI === undefined &&
-      processEnv.CLAUDE_CODE_USE_GITHUB === undefined &&
-      processEnv.CLAUDE_CODE_USE_BEDROCK === undefined &&
-      processEnv.CLAUDE_CODE_USE_VERTEX === undefined &&
-      processEnv.CLAUDE_CODE_USE_FOUNDRY === undefined &&
-      sameOptionalEnvValue(processEnv.MISTRAL_BASE_URL, profile.baseUrl) &&
-      sameOptionalEnvValue(processEnv.MISTRAL_MODEL, profile.model) &&
-      (!includeApiKey ||
-        sameOptionalEnvValue(processEnv.MISTRAL_API_KEY, profile.apiKey))
-    )
-  }
-
-  if (profile.provider === 'gemini') {
-    return (
-      processEnv.CLAUDE_CODE_USE_GEMINI !== undefined &&
-      processEnv.CLAUDE_CODE_USE_MISTRAL === undefined &&
-      processEnv.CLAUDE_CODE_USE_OPENAI === undefined &&
-      processEnv.CLAUDE_CODE_USE_GITHUB === undefined &&
-      processEnv.CLAUDE_CODE_USE_BEDROCK === undefined &&
-      processEnv.CLAUDE_CODE_USE_VERTEX === undefined &&
-      processEnv.CLAUDE_CODE_USE_FOUNDRY === undefined &&
-      sameOptionalEnvValue(processEnv.GEMINI_BASE_URL, profile.baseUrl) &&
-      sameOptionalEnvValue(processEnv.GEMINI_MODEL, profile.model) &&
-      (!includeApiKey ||
-        sameOptionalEnvValue(processEnv.GEMINI_API_KEY, profile.apiKey))
-    )
-  }
-
  return (
    processEnv.CLAUDE_CODE_USE_OPENAI !== undefined &&
    processEnv.CLAUDE_CODE_USE_GEMINI === undefined &&
@@ -415,7 +326,7 @@ function isProcessEnvAlignedWithProfile(
    processEnv.CLAUDE_CODE_USE_VERTEX === undefined &&
    processEnv.CLAUDE_CODE_USE_FOUNDRY === undefined &&
    sameOptionalEnvValue(processEnv.OPENAI_BASE_URL, profile.baseUrl) &&
-    sameOptionalEnvValue(processEnv.OPENAI_MODEL, getPrimaryModel(profile.model)) &&
+    sameOptionalEnvValue(processEnv.OPENAI_MODEL, profile.model) &&
    (!includeApiKey ||
      sameOptionalEnvValue(processEnv.OPENAI_API_KEY, profile.apiKey))
  )
@@ -454,22 +365,6 @@ export function clearProviderProfileEnvFromProcessEnv(
  delete processEnv.ANTHROPIC_API_KEY
  delete processEnv[PROFILE_ENV_APPLIED_FLAG]
  delete processEnv[PROFILE_ENV_APPLIED_ID]
-
-  delete processEnv.GEMINI_MODEL
-  delete processEnv.GEMINI_BASE_URL
-  delete processEnv.GEMINI_API_KEY
-  delete processEnv.GEMINI_AUTH_MODE
-  delete processEnv.GEMINI_ACCESS_TOKEN
-  delete processEnv.GOOGLE_API_KEY
-
-  delete processEnv.MISTRAL_MODEL
-  delete processEnv.MISTRAL_BASE_URL
-  delete processEnv.MISTRAL_API_KEY
-
-  // Clear provider-specific API keys
-  delete processEnv.MINIMAX_API_KEY
-  delete processEnv.NVIDIA_API_KEY
-  delete processEnv.NVIDIA_NIM
 }

 export function applyProviderProfileToProcessEnv(profile: ProviderProfile): void {
@@ -477,7 +372,7 @@ export function applyProviderProfileToProcessEnv(profile: ProviderProfile): void
  process.env[PROFILE_ENV_APPLIED_FLAG] = '1'
  process.env[PROFILE_ENV_APPLIED_ID] = profile.id

-  process.env.ANTHROPIC_MODEL = getPrimaryModel(profile.model)
+  process.env.ANTHROPIC_MODEL = profile.model
  if (profile.provider === 'anthropic') {
    process.env.ANTHROPIC_BASE_URL = profile.baseUrl

@@ -494,54 +389,12 @@ export function applyProviderProfileToProcessEnv(profile: ProviderProfile): void
    return
  }

-  if (profile.provider === 'mistral') {
-    process.env.CLAUDE_CODE_USE_MISTRAL = '1'
-    process.env.MISTRAL_BASE_URL = profile.baseUrl
-    process.env.MISTRAL_MODEL = profile.model
-
-    if (profile.apiKey) {
-      process.env.MISTRAL_API_KEY = profile.apiKey
-    } else {
-      delete process.env.MISTRAL_API_KEY
-    }
-
-    delete process.env.OPENAI_BASE_URL
-    delete process.env.OPENAI_API_KEY
-    delete process.env.OPENAI_MODEL
-    return
-  }
-
-  if (profile.provider === 'gemini') {
-    process.env.CLAUDE_CODE_USE_GEMINI = '1'
-    process.env.GEMINI_BASE_URL = profile.baseUrl
-    process.env.GEMINI_MODEL = profile.model
-
-    if (profile.apiKey) {
-      process.env.GEMINI_API_KEY = profile.apiKey
-    } else {
-      delete process.env.GEMINI_API_KEY
-    }
-
-    delete process.env.OPENAI_BASE_URL
-    delete process.env.OPENAI_API_KEY
-    delete process.env.OPENAI_MODEL
-    return
-  }
-
  process.env.CLAUDE_CODE_USE_OPENAI = '1'
  process.env.OPENAI_BASE_URL = profile.baseUrl
-  process.env.OPENAI_MODEL = getPrimaryModel(profile.model)
+  process.env.OPENAI_MODEL = profile.model

  if (profile.apiKey) {
    process.env.OPENAI_API_KEY = profile.apiKey
-    // Also set provider-specific API keys for detection
-    const baseUrl = profile.baseUrl.toLowerCase()
-    if (baseUrl.includes('minimax')) {
-      process.env.MINIMAX_API_KEY = profile.apiKey
-    }
-    if (baseUrl.includes('nvidia') || baseUrl.includes('integrate.api.nvidia')) {
-      process.env.NVIDIA_API_KEY = profile.apiKey
-    }
  } else {
    delete process.env.OPENAI_API_KEY
  }
@@ -613,7 +466,7 @@ export function addProviderProfile(

  const activeProfile = getActiveProviderProfile()
  if (activeProfile?.id === profile.id) {
-    setActiveProviderProfile(profile.id)
+    applyProviderProfileToProcessEnv(profile)
    clearActiveOpenAIModelOptionsCache()
  }

@@ -695,16 +548,6 @@ export function persistActiveProviderProfileModel(
    return null
  }

-  // If the model is already part of the profile's model list, don't
-  // overwrite the field. This preserves comma-separated model lists like
-  // "glm-4.5, glm-4.7". Switching between models in the list is a
-  // session-level choice handled by mainLoopModelOverride, not a profile
-  // edit — the profile's model list should only change via explicit edit.
-  const existingModels = parseModelList(activeProfile.model)
-  if (existingModels.includes(nextModel)) {
-    return activeProfile
-  }
-
  saveGlobalConfig(current => {
    const currentProfiles = getProviderProfiles(current)
    const profileIndex = currentProfiles.findIndex(
@@ -747,23 +590,6 @@ export function persistActiveProviderProfileModel(
  return resolvedProfile
 }

-/**
- * Generate model options from a provider profile's model field.
- * Each comma-separated model becomes a separate option in the picker.
- */
-export function getProfileModelOptions(profile: ProviderProfile): ModelOption[] {
-  const models = parseModelList(profile.model)
-  if (models.length === 0) {
-    return []
-  }
-
-  return models.map(model => ({
-    value: model,
-    label: model,
-    description: `Provider: ${profile.name}`,
-  }))
-}
-
 export function setActiveProviderProfile(
  profileId: string,
 ): ProviderProfile | null {
@@ -775,85 +601,13 @@ export function setActiveProviderProfile(
    return null
  }

-  const profileModelOptions = getProfileModelOptions(activeProfile)
-
  saveGlobalConfig(config => ({
    ...config,
    activeProviderProfileId: profileId,
-    openaiAdditionalModelOptionsCache: profileModelOptions.length > 0
-      ? profileModelOptions
-      : getModelCacheByProfile(profileId, config),
-    openaiAdditionalModelOptionsCacheByProfile: {
-      ...(config.openaiAdditionalModelOptionsCacheByProfile ?? {}),
-      [profileId]: profileModelOptions.length > 0
-        ? profileModelOptions
-        : (config.openaiAdditionalModelOptionsCacheByProfile?.[profileId] ?? []),
-    },
+    openaiAdditionalModelOptionsCache: getModelCacheByProfile(profileId, config),
  }))

  applyProviderProfileToProcessEnv(activeProfile)
-
-  // Keep startup persisted provider profile in sync so initial startup
-  // uses the selected provider/model.
-  const persistedProfile = (() => {
-    if (activeProfile.provider === 'anthropic') return 'openai' as const
-    return activeProfile.provider
-  })()
-
-  const profileEnv = (() => {
-    switch (activeProfile.provider) {
-      case 'gemini':
-        return (
-          buildGeminiProfileEnv({
-            model: activeProfile.model,
-            baseUrl: activeProfile.baseUrl,
-            apiKey: activeProfile.apiKey,
-            authMode: 'api-key',
-            processEnv: process.env,
-          }) ?? null
-        )
-      case 'mistral':
-        return (
-          buildMistralProfileEnv({
-            model: activeProfile.model,
-            baseUrl: activeProfile.baseUrl,
-            apiKey: activeProfile.apiKey,
-            processEnv: process.env,
-          }) ?? null
-        )
-      default:
-        // anthropic and all openai-compatible providers
-        return (
-          buildOpenAIProfileEnv({
-            model: activeProfile.model,
-            baseUrl: activeProfile.baseUrl,
-            apiKey: activeProfile.apiKey,
-            processEnv: process.env,
-          }) ?? null
-        )
-    }
-  })()
-
-  if (profileEnv) {
-    const startupProfile =
-      activeProfile.provider === 'anthropic'
-        ? ({
-            profile: 'openai' as ProviderProfileStartup,
-            env: {
-              OPENAI_BASE_URL: activeProfile.baseUrl,
-              OPENAI_MODEL: activeProfile.model,
-              OPENAI_API_KEY: activeProfile.apiKey,
-            },
-          } as const)
-        : ({
-            profile: activeProfile.provider as ProviderProfileStartup,
-            env: profileEnv,
-          } as const)
-
-    const file = createProfileFile(startupProfile.profile, startupProfile.env)
-    saveProfileFile(file)
-  }
-
  return activeProfile
 }

--- a/src/utils/providerSecrets.ts
+++ b/src/utils/providerSecrets.ts
@@ -61,7 +61,15 @@ export function maskSecretForDisplay(
    return 'configured'
  }

-  return `${sanitized.slice(0, 3)}...${sanitized.slice(-3)}`
+  if (sanitized.startsWith('sk-')) {
+    return `${sanitized.slice(0, 3)}...${sanitized.slice(-4)}`
+  }
+
+  if (sanitized.startsWith('AIza')) {
+    return `${sanitized.slice(0, 4)}...${sanitized.slice(-4)}`
+  }
+
+  return `${sanitized.slice(0, 2)}...${sanitized.slice(-4)}`
 }

 export function redactSecretValueForDisplay(
--- a/src/utils/providerValidation.test.ts
+++ b/src/utils/providerValidation.test.ts
@@ -1,9 +1,6 @@
 import { afterEach, expect, test } from 'bun:test'

-import {
-  getProviderValidationError,
-  shouldExitForStartupProviderValidationError,
-} from './providerValidation.ts'
+import { getProviderValidationError } from './providerValidation.ts'

 const originalEnv = {
  CLAUDE_CODE_USE_OPENAI: process.env.CLAUDE_CODE_USE_OPENAI,
@@ -96,45 +93,3 @@ test('openai missing key error includes recovery guidance and config locations',
  expect(message).toContain('Saved startup settings can come from')
  expect(message).toContain('.openclaude-profile.json')
 })
-
-test('startup provider validation allows interactive recovery', () => {
-  expect(
-    shouldExitForStartupProviderValidationError({
-      args: [],
-      stdoutIsTTY: true,
-    }),
-  ).toBe(false)
-})
-
-test('startup provider validation stays strict for non-interactive launches', () => {
-  expect(
-    shouldExitForStartupProviderValidationError({
-      args: ['-p', 'hello'],
-      stdoutIsTTY: true,
-    }),
-  ).toBe(true)
-  expect(
-    shouldExitForStartupProviderValidationError({
-      args: ['--print', 'hello'],
-      stdoutIsTTY: true,
-    }),
-  ).toBe(true)
-  expect(
-    shouldExitForStartupProviderValidationError({
-      args: [],
-      stdoutIsTTY: false,
-    }),
-  ).toBe(true)
-  expect(
-    shouldExitForStartupProviderValidationError({
-      args: ['--sdk-url', 'ws://127.0.0.1:3000'],
-      stdoutIsTTY: true,
-    }),
-  ).toBe(true)
-  expect(
-    shouldExitForStartupProviderValidationError({
-      args: ['--sdk-url=ws://127.0.0.1:3000'],
-      stdoutIsTTY: true,
-    }),
-  ).toBe(true)
-})
--- a/src/utils/providerValidation.ts
+++ b/src/utils/providerValidation.ts
@@ -169,44 +169,3 @@ export async function validateProviderEnvOrExit(
    process.exit(1)
  }
 }
-
-export function shouldExitForStartupProviderValidationError(options: {
-  args?: string[]
-  stdoutIsTTY?: boolean
-} = {}): boolean {
-  const args = options.args ?? process.argv.slice(2)
-  const stdoutIsTTY = options.stdoutIsTTY ?? process.stdout.isTTY
-
-  if (!stdoutIsTTY) {
-    return true
-  }
-
-  return (
-    args.includes('-p') ||
-    args.includes('--print') ||
-    args.includes('--init-only') ||
-    args.some(arg => arg.startsWith('--sdk-url'))
-  )
-}
-
-export async function validateProviderEnvForStartupOrExit(
-  env: NodeJS.ProcessEnv = process.env,
-  options?: {
-    args?: string[]
-    stdoutIsTTY?: boolean
-  },
-): Promise<void> {
-  const error = await getProviderValidationError(env)
-  if (!error) {
-    return
-  }
-
-  if (shouldExitForStartupProviderValidationError(options)) {
-    console.error(error)
-    process.exit(1)
-  }
-
-  console.error(
-    `Warning: provider configuration is incomplete.\n${error}\nOpenClaude will continue starting so you can run /provider and repair the saved provider settings.`,
-  )
-}
--- a/src/utils/swarm/spawnUtils.test.ts
+++ b/src/utils/swarm/spawnUtils.test.ts
@@ -1,33 +0,0 @@
-import { afterEach, beforeEach, expect, test } from 'bun:test'
-
-import { buildInheritedEnvVars } from './spawnUtils.js'
-
-const ORIGINAL_ENV = { ...process.env }
-
-beforeEach(() => {
-  for (const key of Object.keys(process.env)) {
-    delete process.env[key]
-  }
-})
-
-afterEach(() => {
-  for (const key of Object.keys(process.env)) {
-    delete process.env[key]
-  }
-  Object.assign(process.env, ORIGINAL_ENV)
-})
-
-test('buildInheritedEnvVars marks spawned teammates as host-managed for provider routing', () => {
-  const envVars = buildInheritedEnvVars()
-
-  expect(envVars).toContain('CLAUDE_CODE_PROVIDER_MANAGED_BY_HOST=1')
-})
-
-test('buildInheritedEnvVars forwards PATH for source-built teammate tool lookups', () => {
-  process.env.PATH = '/custom/bin:/usr/bin'
-
-  const envVars = buildInheritedEnvVars()
-
-  expect(envVars).toContain('PATH=')
-  expect(envVars).toContain('/custom/bin\\:/usr/bin')
-})
--- a/src/utils/swarm/spawnUtils.ts
+++ b/src/utils/swarm/spawnUtils.ts
@@ -141,9 +141,6 @@ const TEAMMATE_ENV_VARS = [
  'NODE_EXTRA_CA_CERTS',
  'REQUESTS_CA_BUNDLE',
  'CURL_CA_BUNDLE',
-  // Source builds may rely on user shell PATH for rg/node/bun and other tools.
-  // Forward it so teammates resolve the same toolchain as the parent session.
-  'PATH',
 ] as const

 /**
@@ -152,13 +149,7 @@ const TEAMMATE_ENV_VARS = [
 * plus any provider/config env vars that are set in the current process.
 */
 export function buildInheritedEnvVars(): string {
-  const envVars = [
-    'CLAUDECODE=1',
-    'CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1',
-    // Teammates should inherit the leader-selected provider route instead of
-    // replaying persisted ~/.claude or settings.env provider defaults.
-    'CLAUDE_CODE_PROVIDER_MANAGED_BY_HOST=1',
-  ]
+  const envVars = ['CLAUDECODE=1', 'CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1']

  for (const key of TEAMMATE_ENV_VARS) {
    const value = process.env[key]
--- a/Show More
+++ b/Show More