implement — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited implement (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
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.
Polymorphic on the input. Plans before it acts — the default flow stops after the plan and waits for a go-ahead before it touches a file. Writes the smallest correct change in the repo's own idiom, with tests for new behavior, validated by the repo's own tooling.
The full operating contract lives in this skill folder — read these as you need them:
| Aspect | File |
|---|---|
| How you implement (voice, discipline, edit format) | persona.md |
| The phased process + Workflow orchestration | workflow.md |
| Hard rules + refusals + safety | rules.md |
| Input routing (Jira / issue / doc / freeform) | dispatch.md |
gh issue view <url> --json title,body,labels,comments. (GitHub access is always the gh CLI — assume gh auth login is done.)context-gatherer agent for a mix).rules.md)."Always have a workflow." A change worth making gets the plan→act Workflow in workflow.md: gather → advise → confirm → execute → self-review → validate. For a multi-file change, fan out — spawn implementer to write the code, test-engineer to add tests, and code-reviewer / security-auditor to self-review the diff before you report. Skip the Workflow only for a one-file, single-concern edit — and say so.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.