| name | platform-health |
| description | Audit platform health — dependency updates, uptime, error rates, CI status, and deployment hygiene. Use for "dependency audit", "check Sentry errors", "platform health review", or scheduled maintenance. Do NOT use for feature implementation (implement) or debugging a single bug (debug).
|
| license | MIT |
| allowed-tools | Read Bash Glob Grep |
| argument-hint | [--focus deps|errors|ci|uptime] |
| metadata | {"author":"Carinya Parc","version":"0.1.0","owner":"web-development","review_cadence":"monthly","work_shape":"monitor-and-report","output_class":"tracking-update"} |
Platform health
Dependency audit, error review, and deployment hygiene for webops workflows.
Optional focus: deps, errors, ci, uptime (default: all).
Steps
- Read target repo
AGENTS.md / CLAUDE.md for commands and conventions.
- Read
.agency/target.json on the target repo, then instance config/ if
present.
- Run checks for the requested focus (or all):
Dependencies (deps)
- Run
pnpm outdated or equivalent from repo root
- Note security advisories (
pnpm audit if available)
- Flag pinned versions with known CVEs or major drift
Errors (errors)
- Review Sentry (if MCP connected) for new/regressed issues in last 7 days
- Group by frequency and user impact; distinguish noise from actionable defects
CI (ci)
- Check recent workflow runs on default branch (GitHub MCP or
gh run list)
- Note flaky or failing jobs; verify required checks match branch protection
Uptime (uptime)
- Check Vercel deployment status (if MCP connected)
- Note failed or rolled-back production deploys in last 14 days
- Confirm preview deployments succeed for open PRs when relevant
Constraints
- Do not modify application code or merge PRs
- File recommendations as GitHub issues when actionable
- Escalate production incidents to a human immediately
Output format
## Platform health — {date}
### Summary
(one paragraph — overall status: green / amber / red)
### Findings
| Area | Severity | Finding | Recommended action |
| ---- | -------- | ------- | ------------------ |
### Suggested issues
- [ ] {title} — squad:site, type:maintenance
Severity: critical · high · medium · low · info.