openplan-review — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited openplan-review (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Use this after $openplan-core to review automated agent work without becoming another ungoverned agent. The review must be evidence-backed, read-only by default, preserve dual-surface separation when present, and remain independent of any one platform's enforcement implementation.
Do not approve, merge, commit, rewrite, or fix the work under review. Inspect, verify, classify, and report. If the user explicitly asks for fixes after the review, treat that as a separate task.
references/invariants.md.references/output-contract.md.python3 scripts/validate_review_report.py <report.md>BLOCKED with the exact missing artifact.git reset --hard, branch checkout, deleting working directories, cleanup of untracked files).Non-Decisions.P0: false approval risk, data loss, security leak, broken authority flow, or review cannot establish what changed.P1: architectural/workflow contradiction, human decision bypass, mode/role boundary bypass, research skipped where required, generated workflow divergence.P2: test gap, stale reference, incomplete traceability, unclear ownership, missing evidence.P3: style, naming, wording, or maintainability issue with low behavioral risk.$openplan-review.assets/claude-code-command.md as a slash command or reviewer-agent prompt.$openplan-handoff so the next agent gets a minimal snapshot instead of a transcript.references/invariants.md: OpenPlan philosophy and review checklist.references/output-contract.md: required report shape.references/platform-usage.md: Codex and Claude Code integration notes.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.