Telegram (ask questions / claim the issue here first): https://t.me/+DOylgFv1jyJlNzM0
Why this matters
docs/runbooks/ contains only indexer-recovery.md, yet deploy-staging.yml implements a non-trivial pipeline: staging-previous backup/rollback, vars.STAGING_ENABLED == true gating, required secrets (STAGING_SSH_HOST/USER/KEY/PORT), server-side migrate:up, and backend/frontend health checks. There is no runbook telling an on-call engineer how to enable staging, which secrets/vars to set, or how to manually roll back to remitlend-*:staging-previous when the automated rollback also fails.
Acceptance criteria
Files to touch
docs/runbooks/README.md
.github/workflows/deploy-staging.yml
docker-compose.staging.yml
Out of scope
- Documenting a production deploy pipeline that does not yet exist
- Secret rotation policy
Why this matters
docs/runbooks/ contains only indexer-recovery.md, yet deploy-staging.yml implements a non-trivial pipeline: staging-previous backup/rollback, vars.STAGING_ENABLED == true gating, required secrets (STAGING_SSH_HOST/USER/KEY/PORT), server-side migrate:up, and backend/frontend health checks. There is no runbook telling an on-call engineer how to enable staging, which secrets/vars to set, or how to manually roll back to remitlend-*:staging-previous when the automated rollback also fails.
Acceptance criteria
Files to touch
docs/runbooks/README.md.github/workflows/deploy-staging.ymldocker-compose.staging.ymlOut of scope