feat(cache): Anthropic prompt caching, usage telemetry, and Cache Rate view

- send prompt-caching + oauth anthropic-beta headers on the Claude OAuth provider
- restructure the OAuth request body (billing header, identity split, relocate
  third-party system prompt to the first user message) to match Claude Code
- apply rolling cache_control breakpoints and group a turn's tool results into a
  single role:tool message for correct breakpoint placement
- emit per-step usage events (cache read/write split) and add the Cache Rate
  sidebar panel
- dedup byte-identical tool calls within a single batch

1 files changed, 5 insertions, 1 deletions

diff --git a/packages/core/tests/tools/summon.test.ts b/packages/core/tests/tools/summon.test.ts
index 3909e48..722748c 100644
--- a/packages/core/tests/tools/summon.test.ts
+++ b/packages/core/tests/tools/summon.test.ts
@@ -118,7 +118,11 @@ describe("createSummonTool — execute() argument forwarding", () => {
 			[],
 		);
 		const out = await tool.execute({ task: "x" });
-		expect(out).toBe("child-output");
+		// Foreground summons prefix the blocked result with `agent_id: <id>` so
+		// the frontend's ToolCallDisplay regex can surface the "Open Tab" button
+		// (see summon.ts). Assert both the prefix and the child output survive.
+		expect(out).toContain("agent_id: id-1");
+		expect(out).toBe("agent_id: id-1\n\nchild-output");
 	});
 
 	it("surfaces child errors when blocking", async () => {