$15 one-time. Five Claude Code skills + one subagent for the judgment calls generic AI assistance gets wrong: migration safety, real changelogs, blameless postmortems, dependency-upgrade risk, flaky-test diagnosis.
db-migration-guard — lock behavior, rolling-deploy compatibility, backfill/rollback safety reviewapi-changelog — commit range → categorized, consumer-facing changelogincident-postmortem — raw notes/logs → timeline + real root cause + blameless writeupdependency-upgrade-planner — ranks breaking-change risk by what your code actually callsflaky-test-detector — diagnoses timing/shared-state/ordering/external-dep mechanismsrelease-manager subagent — end-to-end release drafting, read-only by defaultPlain markdown, no build step, no telemetry, no dependencies. See the free guides on diagnosing flaky tests for the checklist behind flaky-test-detector, on safe database migrations for the checklist behind db-migration-guard, on dependency upgrade risk for the checklist behind dependency-upgrade-planner, and on blameless postmortems for the checklist behind incident-postmortem.
0xa7B251CA5306d177ED2D00f181121D2830c47fD8Automated check not working? Email the tx hash to orders@intirama.com with subject "Power Pack" for a manual check.