Skip to content

feat(reliability): add production runtime canary policy - #452

Merged
Yevanchen merged 1 commit into
mainfrom
feat/reliability-status-canary
Jul 29, 2026
Merged

feat(reliability): add production runtime canary policy#452
Yevanchen merged 1 commit into
mainfrom
feat/reliability-status-canary

Conversation

@Yevanchen

Copy link
Copy Markdown
Collaborator

Summary

  • Add a secret-protected internal diagnostic that verifies two canary Runs stayed on the same driver without exposing driver IDs publicly.
  • Reuse the existing constant-time internal-secret comparison for the Lark sidecar and status canary routes.
  • Document the five-minute production canary, 99.5% SLO, three-breach release freeze, and one-page incident postmortem discipline.

Why

Production regressions such as #431 can pass staging but break follow-up driver continuity. This gives the public website canary one narrow, authenticated signal for that production-only failure mode.

Validation

  • just test-file apps/api/tests/status-canary-internal-route.test.ts
  • just tc-package @mosoo/api
  • just lint
  • just docs-check
  • just commit-check
  • just check reached 1080 passing tests and failed only the unrelated existing macOS /var vs /private/var assertion in apps/driver/tests/acp-file-system.test.ts; the focused failure reproduces without touching this PR's API path.

Generated and migration state

  • GraphQL codegen: not applicable
  • D1 migrations: none
  • Generated files: none
  • Lockfile changes: none

@Yevanchen
Yevanchen merged commit 4a58e09 into main Jul 29, 2026
9 checks passed
@Yevanchen
Yevanchen deleted the feat/reliability-status-canary branch July 29, 2026 16:11
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant