docs/current-handoff.md

@@ -98,6 +98,12 @@
 - This skipped schema-constrained reconstruction `/api/chat` and used unconstrained `/api/generate`. Response-body disposal now uses Fetch-compatible consumption without changing capability or fallback semantics.
 - Next boundary: one fresh-process fixed-scenario production observation.
 
+## Reconstruction chat timeout
+
+- Capability detection/body disposal was fixed at `7070342`. Fresh runtime evidence then proved the real schema-constrained `/api/chat` request was attempted but Confidence Engine aborted it after 60 seconds while Qwen was still generating.
+- Fallback `/api/generate` completed but produced structurally invalid reconstruction. The real reconstruction `/api/chat` timeout is now 300 seconds, matching `/api/generate`; fallback, schema, prompt, and reasoning semantics are unchanged.
+- Next boundary: one fresh-process fixed-scenario production observation.
+
 ## Current product architecture
 
 Three distinct routes, not a single page:

lib/llm/provider.js

@@ -134,7 +134,7 @@ class OllamaLlmProvider {
     if (chatSupported) {
       try {
         const controller = new AbortController();
-        const timeout = setTimeout(() => controller.abort(), 60000);
+        const timeout = setTimeout(() => controller.abort(), 300000); // 5 min for reconstruction
 
         apiUsed = "/api/chat";
         providerExecution.chatRequestAttempted = true;

tests/llm/provider.test.js

@@ -10,6 +10,7 @@ describe("OllamaLlmProvider chat capability detection", () => {
         ok: true,
         json: async () => ({ message: { content: "{}" } }),
       });
+    const setTimeoutSpy = vi.spyOn(globalThis, "setTimeout");
     vi.stubGlobal("fetch", fetchSpy);
     process.env.OLLAMA_BASE_URL = "http://ollama.test";
 
@@ -24,6 +25,7 @@ describe("OllamaLlmProvider chat capability detection", () => {
       expect(fetchSpy.mock.calls[0][0]).toBe("http://ollama.test/api/chat");
       expect(fetchSpy.mock.calls[1][0]).toBe("http://ollama.test/api/chat");
       expect(fetchSpy.mock.calls[1][0]).not.toContain("/api/generate");
+      expect(setTimeoutSpy).toHaveBeenCalledWith(expect.any(Function), 300000);
       const chatRequest = JSON.parse(fetchSpy.mock.calls[1][1].body);
       expect(chatRequest).toMatchObject({
         model: "configured-model",
@@ -46,6 +48,7 @@ describe("OllamaLlmProvider chat capability detection", () => {
         },
       });
     } finally {
+      setTimeoutSpy.mockRestore();
       vi.unstubAllGlobals();
       if (originalBaseUrl === undefined) delete process.env.OLLAMA_BASE_URL;
       else process.env.OLLAMA_BASE_URL = originalBaseUrl;