Benchmark report

Task verified

7m 31s elapsed

The saved attempt met the documented success criterion.

Experience score

93/ 100

Agent-Ready
  • AInstall
  • SAuth
  • SExecute
  • ADocs
  • ACurrent
Experience measures friction in the tested task. The verdict says whether it succeeded.
Agent profile for this run: Install A, Auth S, Execute S, Docs A, Current AAInstallSAuthSExecuteADocsACurrent
Agent profile · this run
Judge's note · what slowed the agent most
The PEP 668 pip failure plus one 404 from guessing a 'stop-browser-session' doc page before llms.txt revealed PATCH update-browser-session.

Where the time went

7:31 agent time
  • Research0:32
  • Attempt4:54
  • Verify2:04
7m 31sTotal elapsed
3:17Time to first success
0:00Waited on the owner

Judge reason

Saved code (evidence/run1_code_cookie_carry.py, identical to work/cookie_carry.py) and live responses show the full documented v4 flow: POST /api/v4/profiles -> 201 id e499780c-a5ca-4585-8b96-19d5632a4118; POST /api/v4/browsers {profileId} -> 201 session A 6cdee3b6...f2ef; Playwright connect_over_cdp to the returned cdpUrl https://6cdee3b6....cdp.browser-use.com (no key in URL/query) loaded httpbin.org/cookies/set?tpc=benchmark, page body {"cookies":{"tpc":"benchmark"}}; PATCH /browsers/{A} {"action":"stop"} -> 200 status=stopped finishedAt 23:09:10.938Z, polled to stopped, GET /profiles -> cookieDomains ["httpbin.org"]; session B created afterwards (startedAt 23:09:11.427Z) with the same profileId, distinct id 1c0e922a...4f7f, and its httpbin.org/cookies response was {"cookies":{"tpc":"benchmark"}} (run1_sessionB_page.json, run1_09_RESULT_cookie_carry.json). finally block: PATCH stop B -> 200 stopped, DELETE /profiles/{id} -> 204. All documented operations match the fetched doc snapshots (playwright-puppeteer-selenium, update-browser-session action=stop, delete-browser-profile 204). Console log matches the saved JSON responses; nothing mocked (persist_cookie_flag false, so the add_cookies fallback branch never ran); the key appears only as X-Browser-Use-API-Key header and no saved artifact contains a bu_ key on independent grep.

Attempt record

  1. 2:29Answered from saved key: Browser Use Cloud API keyno wait
  2. —First success verified

Attempt artifacts saved.