feat(cli): profile ingest runs and split model vs tool time (#249)

* feat(cli): profile ingest runs to find where wall-clock time goes

Add opt-in profiling for `ktx ingest`. Each timed phase, work unit, and
agent loop now records durationMs / step count / token usage in the
trace, and a post-run aggregator rolls them up into a "where did the
time go" report printed to stderr.

Enable per run with KTX_PROFILE_INGEST (1/true -> human table, json ->
raw structured profile) or persistently via `ingest.profile` in
ktx.yaml. The json form emits raw milliseconds, token counts, and a
summary.headline one-line diagnosis so coding agents can parse it
directly; json wins when both env and config request profiling.

- runtime-port: RunLoopMetrics (totalMs, usage, stepCount,
  stepBoundariesMs) plus onMetrics callbacks on text/object generation
- ai-sdk + claude-code runtimes: capture per-loop timing and token usage
- work-unit-executor and stages 3/4: thread metrics into trace events
- ingest-bundle.runner: time worktree / triage / clustering / index /
  reconcile / squash phases and emit the profile in a finally block
  (best-effort; never affects the run outcome)
- ingest-profile: new trace+transcript aggregator with table/json formatters
- config: ingest.profile flag; docs: profiling section in ktx-ingest.mdx

* fix(cli): flush tool-call logs before reading ingest profile

Tool transcripts are appended fire-and-forget so the agent hot path never
blocks on logging. The ingest profiler read them before the writes settled,
so per-work-unit toolMs (and the model-vs-tool split derived from it) could
be incomplete. Track in-flight appends and expose flushToolCallLogs() —
bounded by a timeout so it can never hang — and flush before the profiler
reads the transcript.

packages/cli/test/context/llm/claude-code-runtime.test.ts

@@ -284,7 +284,7 @@ describe('ClaudeCodeKtxLlmRuntime', () => {
         stepBudget: 1,
         telemetryTags: { operationName: 'test' },
       }),
-    ).resolves.toEqual({ stopReason: 'budget' });
+    ).resolves.toMatchObject({ stopReason: 'budget' });
 
     const options = query.mock.calls[0][0].options;
     expect(options.allowedTools).toEqual(['mcp__ktx__load_skill']);
@@ -467,7 +467,7 @@ describe('ClaudeCodeKtxLlmRuntime', () => {
         telemetryTags: { operationName: 'test' },
         onStepFinish,
       }),
-    ).resolves.toEqual({ stopReason: 'natural' });
+    ).resolves.toMatchObject({ stopReason: 'natural' });
 
     expect(onStepFinish).toHaveBeenCalledTimes(1);
     expect(onStepFinish).toHaveBeenCalledWith({ stepIndex: 1, stepBudget: 40 });
@@ -513,7 +513,7 @@ describe('ClaudeCodeKtxLlmRuntime', () => {
           throw new Error('callback exploded');
         },
       }),
-    ).resolves.toEqual({ stopReason: 'natural' });
+    ).resolves.toMatchObject({ stopReason: 'natural' });
     expect(logger.warn).toHaveBeenCalledWith(expect.stringContaining('callback exploded'));
   });
 
@@ -525,6 +525,45 @@ describe('ClaudeCodeKtxLlmRuntime', () => {
     expect(mapClaudeCodeStopReason(resultMessage({ subtype: 'error_during_execution' }))).toBe('error');
   });
 
+  it('returns loop metrics including step count and mapped token usage', async () => {
+    const query = vi.fn((_input: any) =>
+      stream([
+        initMessage(),
+        {
+          type: 'assistant',
+          message: { role: 'assistant', content: [] },
+          parent_tool_use_id: null,
+          uuid: '00000000-0000-4000-8000-000000000006',
+          session_id: 'session-id',
+        } as unknown as SDKMessage,
+        resultMessage({
+          subtype: 'success',
+          terminal_reason: 'completed',
+          usage: { input_tokens: 50, output_tokens: 10 } as never,
+        }),
+      ]),
+    );
+    const runtime = new ClaudeCodeKtxLlmRuntime({
+      projectDir: '/tmp/project',
+      modelSlots: { default: 'sonnet' },
+      query,
+      env: {},
+    });
+
+    const result = await runtime.runAgentLoop({
+      modelRole: 'default',
+      systemPrompt: 'system',
+      userPrompt: 'user',
+      toolSet: {},
+      stepBudget: 40,
+      telemetryTags: { operationName: 'test' },
+    });
+
+    expect(result.metrics?.stepCount).toBe(1);
+    expect(result.metrics?.stepBoundariesMs).toHaveLength(1);
+    expect(result.metrics?.usage).toEqual({ inputTokens: 50, outputTokens: 10, totalTokens: 60 });
+  });
+
   it('auth probe uses isolation options and a scrubbed env', async () => {
     const query = vi.fn((_input: any) => stream([initMessage(), resultMessage({ result: 'ok' })]));