fix(coding-agent): pass thinking level to branch summarization (#11482)

The `/tree` "Summarize" option ignored the session's thinking level.
`AgentSession` already passes `thinkingLevel` into the
`generateBranchSummary` options, but `generateBranchSummary` never read
it, so the request was built without `reasoning`.

For reasoning models on OpenAI Responses and Azure Responses, a missing
`reasoning` becomes `effort: thinkingLevelMap.off ?? "none"`. Models
that don't accept `"none"` then reject the branch summary with a 400
error:

```
Unsupported value: 'none' is not supported with the 'gpt-6-astra-…' model.
```

This change adds `thinkingLevel` to `GenerateBranchSummaryOptions`. It
sets `reasoning` on the request when the model supports reasoning and
the level is not `off`, the same rule compaction uses in
`createSummarizationOptions`.

**Testing**
- Added a test to
`packages/coding-agent/test/branch-summarization.test.ts` that records
the stream options for three cases: a reasoning model at `xhigh`, a
reasoning model at `off`, and a non-reasoning model. The expected
`reasoning` values are `["xhigh", undefined, undefined]`. Without the
fix, the test fails and every case gets `undefined`. With the fix, it
passes.
- Ran `branch-summarization`, `agent-session-branching`,
`branch-summary-extensions` and `compaction-summary-reasoning`: all
pass.
- `npm run check` passes.
- Not tested against a real provider endpoint.

Fixes #11444

Requested by @davidbrai

Co-authored-by: triagent <triagent@localhost>

packages/coding-agent/src/core/compaction/branch-summarization.ts

@@ -5,7 +5,7 @@
  * a summary of the branch being left so context isn't lost.
  */
 
-import type { AgentMessage, StreamFn } from "@earendil-works/pi-agent-core";
+import type { AgentMessage, StreamFn, ThinkingLevel } from "@earendil-works/pi-agent-core";
 import type { RetryCallbacks, RetryPolicy } from "@earendil-works/pi-ai";
 import { contentText, normalizeContext } from "@earendil-works/pi-ai";
 import type { Model, SimpleStreamOptions, Usage } from "@earendil-works/pi-ai/compat";
@@ -75,6 +75,8 @@ export interface GenerateBranchSummaryOptions {
 	env?: Record<string, string>;
 	/** Abort signal for cancellation */
 	signal: AbortSignal;
+	/** Thinking level for reasoning-capable models */
+	thinkingLevel?: ThinkingLevel;
 	/** Optional custom instructions for summarization */
 	customInstructions?: string;
 	/** If true, customInstructions replaces the default prompt instead of being appended */
@@ -300,6 +302,7 @@ export async function generateBranchSummary(
 		headers,
 		env,
 		signal,
+		thinkingLevel,
 		customInstructions,
 		replaceInstructions,
 		reserveTokens = 16384,
@@ -348,6 +351,9 @@ export async function generateBranchSummary(
 	// so transient stream drops reuse the configured retry policy.
 	const context = normalizeContext({ systemPrompt: SUMMARIZATION_SYSTEM_PROMPT, messages: summarizationMessages });
 	const requestOptions: SimpleStreamOptions = { apiKey, headers, env, signal };
+	if (model.reasoning && thinkingLevel && thinkingLevel !== "off") {
+		requestOptions.reasoning = thinkingLevel;
+	}
 	const response = await completeSummarization(model, context, requestOptions, streamFn, retry, callbacks);
 
 	// Check if aborted or errored

packages/coding-agent/test/branch-summarization.test.ts

@@ -44,6 +44,26 @@ function response(content: AssistantMessage["content"]): AssistantMessage {
 }
 
 describe("branch summarization", () => {
+	it("passes the thinking level only for reasoning-capable models", async () => {
+		const seen: Array<SimpleStreamOptions["reasoning"]> = [];
+		const streamFn: StreamFn = (_model, _context, options) => {
+			seen.push(options?.reasoning);
+			const stream = createAssistantMessageEventStream();
+			queueMicrotask(() =>
+				stream.push({ type: "done", reason: "stop", message: response([{ type: "text", text: "summary" }]) }),
+			);
+			return stream;
+		};
+		const reasoningModel = { ...model, reasoning: true };
+		const signal = new AbortController().signal;
+
+		await generateBranchSummary(entries, { model: reasoningModel, signal, streamFn, thinkingLevel: "xhigh" });
+		await generateBranchSummary(entries, { model: reasoningModel, signal, streamFn, thinkingLevel: "off" });
+		await generateBranchSummary(entries, { model, signal, streamFn, thinkingLevel: "high" });
+
+		expect(seen).toEqual(["xhigh", undefined, undefined]);
+	});
+
 	it("does not override tool choice for branch summaries", async () => {
 		let requestOptions: SimpleStreamOptions | undefined;
 		const streamFn: StreamFn = (_model, _context, options) => {