Audits release readiness by classifying risk level, checking verification evidence, and confirming operational readiness before deploying or merging.
How this skill is triggered — by the user, by Claude, or both
Slash command
/agent-coding-playbook:acp-release-checkThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill before releasing, deploying, publishing, or merging a change into a production-bound branch.
Use this skill before releasing, deploying, publishing, or merging a change into a production-bound branch.
The goal is to reduce release risk, not to create a long ceremony.
Summarize:
Classify the release as low, medium, or high risk.
High-risk signals:
Look for:
If evidence is missing, mark it as a release risk.
Confirm:
Release scope:
- ...
Risk level:
- Low / Medium / High
Verification evidence:
- ...
Missing evidence:
- ...
Rollback plan:
- ...
Recommendation:
- Go / Go with caution / No-go
Do not say a release is safe only because the code compiles.
A release is safe when the change scope, verification evidence, runtime observability, and rollback path are all clear.
npx claudepluginhub bravekingzhang/agent-coding-playbook --plugin agent-coding-playbookRun a read-only pre-release review for deploy readiness, migrations, config, secrets, rollout order, rollback risk, and launch blockers.
Validates pre-release readiness by reviewing diffs, changelogs, CI status, tests, and artifacts. Guides versioning, release notes, smoke tests, and rollback planning.
Audits production readiness using local evidence: runtime, auth, data, payments, background tasks, AI, and deployment boundaries. For pre-release, post-merge, or 'what will break in production?' questions.