fix(antigravity): retry transient upstream failures

Retry short-lived 5xx/capacity errors (500/502/503/504 + message
patterns) with bounded backoff capped at 15s; honor Retry-After/reset
hints and skip when wait is too long. Keep 400 non-retryable. Enable the
retry hook for 500 alongside existing 429/503.

Deduplicate sanitized Antigravity tool names before emitting the single
functionDeclarations group to avoid upstream "Tool names must be unique"
rejections.

Add Headroom size diagnostics and phantom-savings warning when reported
token delta does not shrink the outbound payload.

Co-authored-by: Cursor <cursoragent@cursor.com>

tests/unit/base-executor-retry.test.js

@@ -85,6 +85,16 @@ describe("BaseExecutor.execute — network error retry/fallback", () => {
 });
 
 describe("BaseExecutor.execute — computeRetryDelay hook veto", () => {
+  it("only invokes computeRetryDelay when status has retry config", async () => {
+    const ex = makeExec({ baseUrl: "https://x/api", retry: { 503: { attempts: 1, delayMs: 0 } } });
+    ex.computeRetryDelay = vi.fn().mockResolvedValue(0);
+    fetchMock.mockResolvedValueOnce(res(500));
+    const out = await ex.execute({ model: "m", body: {}, stream: false, credentials: creds });
+    expect(out.response.status).toBe(500);
+    expect(ex.computeRetryDelay).not.toHaveBeenCalled();
+    expect(fetchMock).toHaveBeenCalledTimes(1);
+  });
+
   it("hook returning false skips retry (uses fallback path)", async () => {
     const ex = makeExec({ baseUrl: "https://x/api", retry: { 429: { attempts: 5, delayMs: 0 } } });
     ex.computeRetryDelay = vi.fn().mockResolvedValue(false);