OpenAI Codex's run of the AI Travel Planner

A plain Worker (no framework) with D1 and a React 19 front end, branded "Elsewhere". It has a Discover home screen, structured trip cards with optimistic concurrency, and a separate admin app with drill-downs. It committed the most evidence of the three runs, including screenshots and JSON check logs.

Run length Agent claim: Reported by the agent in its own build report. Not independently verified.

1h 28m

of 8h allowed · started 16:36 UTC, Nov 5, 2026

Requirements tracked Agent claim: Reported by the agent in its own build report. Not independently verified.

55

54 marked Verified by the agent

Checks reported passing Agent claim: Reported by the agent in its own build report. Not independently verified.

117

across 6 suites of different scope

Approvals Observed: Observed by The Build Hypothesis during the recorded run.

Asked for approval whenever it needed to run a new command. Commands it had already been allowed could be allowed again.

Timeline

  • Commit
  • Problem
  • Fix
  • Deploy
  • Start / end
View the timeline as a table

OpenAI Codex, started 16:36 UTC

OffsetTimeEventTypeSource
+0 min16:36 UTCRun startedStartAgent claim: Reported by the agent in its own build report. Not independently verified.
+33 min17:08 UTCImplement Elsewhere travel planner with secure Cloudflare backend and reportingCommitGit record: Taken from the public repository's commit history.
+33 min17:09 UTC ≈First Workers Build failed at npm ci (lockfile missing a peer entry)ProblemAgent claim: Reported by the agent in its own build report. Not independently verified.
+42 min17:17 UTCFix Cloudflare npm lock compatibility and verify responsive travel workspaceFixGit record: Taken from the public repository's commit history.
+44 min17:20 UTCFirst production smoke test returned HTTP 500 on admin login (PBKDF2 iteration limit)ProblemAgent claim: Reported by the agent in its own build report. Not independently verified.
+47 min17:22 UTCRetain full-strength password hashing on Cloudflare production runtimeFixGit record: Taken from the public repository's commit history.
+1h 4m17:40 UTCComplete production verification, reporting drill-downs and build handoffCommitGit record: Taken from the public repository's commit history.
+1h 10m17:45 UTCApply security headers to Cloudflare-served HTMLFixGit record: Taken from the public repository's commit history.
+1h 17m17:52 UTCAccount for deleted-trip usage in unattributed costsFixGit record: Taken from the public repository's commit history.
+1h 28m18:03 UTCBuild results generatedEndAgent claim: Reported by the agent in its own build report. Not independently verified.
+1h 28m18:04 UTCPublish final build reports and verified deployment handoffCommitGit record: Taken from the public repository's commit history.

5,262 seconds, measured when the report was generated. The end time was left empty on purpose while publishing was still in progress.

Tests the agent reported

  • 39 passed · 0 failedWorker API tests (Vitest in workerd, local D1, mocked externals)1 file
  • 8 passed · 0 failedLocal browser tests (Playwright, desktop and mobile)
  • 8 passed · 0 failedProduction browser tests
  • 16 passed · 0 failedProduction admin browser checks
  • 22 passed · 0 failedProduction smoke checks (real OpenAI, Google and email binding)
  • 24 passed · 0 failedProduction bootstrap smoke

The agent says "Verified" often means a local test with mocked services, not production verification.

Deployment & cost

Deployment Agent claim: Reported by the agent in its own build report. Not independently verified.
Deployed through the GitHub-connected Workers Builds pipeline. Of 6 push builds during the run, 1 failed and 5 succeeded.
Interventions Agent claim: Reported by the agent in its own build report. Not independently verified.
Counts tool approvals as interventions (dependency installs, Cloudflare provisioning and migrations, test runtime). No manual code changes.
Estimated API spend while testing in production Agent claim: Reported by the agent in its own build report. Not independently verified.
$0.066Estimated for the agent's own production checks (4 LLM calls and 2 Places calls).
Public demo
The experiment deployment calls billable AI and Places APIs, so it is not linked publicly. A demo link will be added here if a public version is published.

Highlights Agent claim: Reported by the agent in its own build report. Not independently verified.

  • Optimistic concurrency on trips (revision numbers), with overlapping schedules rejected
  • The server preserves unrelated days during single-day conversational edits
  • Configurable pricing with stored price snapshots, and unknown prices recorded as null instead of guessed
  • Kept full-strength 600,000-iteration password hashing on Cloudflare by switching libraries rather than lowering the work factor
  • Found that static HTML was missing security headers and fixed it with a _headers file, then re-checked

Gaps it admitted Agent claim: Reported by the agent in its own build report. Not independently verified.

  • Production end-to-end verification is Partial (inbox receipt, link consumption and multi-device access not run)
  • Real inbox receipt of verification and reset emails was blocked, because there was no controlled mailbox
  • Map lines show order, not routed directions; destination photos are decorative stock images
  • No automatic retries on LLM calls, by design, to avoid duplicate billing

Stack & Cloudflare services

Stack

  • TypeScript
  • Cloudflare Workers (no framework)
  • Zod
  • React 19
  • Vite 7
  • Leaflet + OpenStreetMap
  • @noble/hashes
  • Vitest
  • Playwright

Cloudflare

  • Workers
  • D1
  • Workers Static Assets
  • Email Service
  • Workers Builds
  • Observability

Screenshots

Notes

Notes from the build report

  • Sandbox network and socket restrictions meant dependency downloads, Cloudflare operations and workerd tests needed approved retries. Codex counts these tool approvals as human interventions.
  • Production Cloudflare caps native PBKDF2 at 100,000 iterations, but local workerd didn’t, so the first production admin login failed. The agent switched to @noble/hashes at 600,000 iterations instead of weakening the hash.
  • The report keeps its failed checks (the first build and the first production auth check) next to the passing ones.
  • Two synthetic guest trips from the first production browser run remain in the production database, according to the report.

The other runs