Runbooks & On-Call¶
Home / Guides & Runbooks / Runbooks & On-Call
Step-by-step procedures for recurring operational tasks and incident response. A good runbook is something a teammate can follow at 2 a.m. without context — keep each one concrete, sequential, and tested.
Scaffold — add runbooks as procedures are formalized
Start a new ## Runbook: … section for each procedure. When a runbook gets long, split it into its own page under guides/.
On-Call¶
- TBD: rotation, escalation path, how to acknowledge and hand off.
Runbook: Deploy the Docs Site¶
The wiki deploys automatically — pushing to main triggers a Cloudflare Pages build.
- Merge your change to
main. - Watch the Cloudflare Pages build for the
docsproject. - Verify the live site once the build is green.
Runbook: Rotate a Leaked Credential¶
- Identify the affected system in the Systems Registry.
- Revoke the exposed key/token in that system.
- Issue a replacement and update every consumer (env vars, GitHub Actions secrets,
~/.mds/config.toml, etc.). - Note the rotation and confirm the old credential no longer works.
Runbook: Incident Response¶
- TBD: declare, communicate, mitigate, write up. Record postmortems as ADRs in Engineering → Decisions.