fix(api): consolidate 3P provider compatibility fixes

- Strip store field from request body for local providers (Ollama, vLLM)
  that reject unknown JSON fields with 400 errors
- Add Gemini 3.x model context windows and output token limits
  (gemini-3-flash-preview, gemini-3.1-pro-preview, google/ OpenRouter variants)
- Preserve reasoning_content on assistant tool-call message replays
  for providers that require it (Kimi k2.5, DeepSeek reasoner)
- Use conservative max_output_tokens fallback (4096/16384) for unknown
  3P models to prevent vLLM/Ollama 400 errors from exceeding max_model_len

Consolidates fixes from: #258, #268, #237, #643, #666, #677

Co-authored-by: auriti <auriti@users.noreply.github.com>
Co-authored-by: Gustavo-Falci <Gustavo-Falci@users.noreply.github.com>
Co-authored-by: lttlin <lttlin@users.noreply.github.com>
Co-authored-by: Durannd <Durannd@users.noreply.github.com>

src/__tests__/bugfixes.test.ts

@@ -21,11 +21,11 @@ describe('Gemini store field fix', () => {
   test('isGeminiMode is imported and used in openaiShim', async () => {
     const content = await file('services/api/openaiShim.ts').text()
 
-    // Verify the fix: store deletion should check for Gemini mode
+    // Verify the fix: store deletion should check for Gemini mode and local providers
     expect(content).toContain('isGeminiMode()')
-    expect(content).toContain("mistral and gemini don't recognize body.store")
-    // Ensure the delete body.store is guarded for both Mistral and Gemini
-    expect(content).toMatch(/isMistral\s*\|\|\s*isGeminiMode\(\)/)
+    expect(content).toContain("Strip store for providers that don't recognize it")
+    // Ensure the delete body.store is guarded for Mistral, Gemini, and local providers
+    expect(content).toMatch(/isMistral\s*\|\|\s*isGeminiMode\(\)\s*\|\|\s*isLocal/)
   })
 
   test('store: false is still set by default (OpenAI needs it)', async () => {