The Prompting Company

Benchmark report

Onboarding verified

3m 35s elapsed

The saved attempt met the documented success criterion.

Experience score

100/ 100

Agent-Ready
  • SInstall
  • SAuth
  • SExecute
  • SDocs
  • SCurrent
Experience measures friction in the tested task. The verdict says whether it succeeded.
Agent profile for this run: Install S, Auth S, Execute S, Docs S, Current SSInstallSAuthSExecuteSDocsSCurrent
Agent profile · this run
Judge's note · what slowed the agent most
Only the human-in-the-loop wait for the API key (~57s heartbeat on benchmark-request); the product itself added no friction.

Where the time went

0:57 on the owner · 2:39 agent time
Waited on the owner 0:57
  • Starting0:01
  • Research0:18
  • Attempt1:26
  • Verify0:54
3m 35sTotal elapsed
1:39Time to first success
0:57Waited on the owner

Judge reason

Live API call verified, not mocked. Transcript step 23 shows a real curl POST to https://api.context.dev/v1/web/scrape with a Bearer key loaded from the benchmark-request response file; curl exit 0 and HTTP 200 (scrape_http_status.txt = 200). scrape_headers.txt shows genuine server response headers (HTTP/2 200, cf-ray, server: cloudflare, x-credits-used: 1, x-request-id: bcf64def-3cc6-4a79-a886-d27a8f18daab) whose request_id matches the body in scrape_result.json. The saved response satisfies every criterion: markdown.requested=true, markdown.data is non-empty and begins '# Example Domain', metadata.title='Example Domain', key_metadata.credits_consumed=1. Step 25-26 independently re-parsed the file with jq confirming the same values. No API key material appears in any artifact (grep for ctxt_secret/Bearer tokens finds only the env-var name in result.json); the agent even stripped a 'ctxt_secret_***REDACTED***' placeholder from command_run.txt at step 39. An unauthenticated probe at step 8 returned HTTP 401, further showing auth was genuinely enforced.

Attempt record

  1. 0:49Asked the owner for Context.dev API key (CONTEXT_DEV_API_KEY, format ctxt_secret_...)waited 57s · provided
  2. —First success verified

Attempt artifacts saved.