_template — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited _template (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.
Use this file as the starting point for a new techne skill. Keep the skill body tool-neutral: the guidance here should make sense outside any one host, with tool-specific loading handled by plugin manifests.
A techne skill must force a move the model can make but often skips. It is not a place to restate common knowledge, collect vague advice, or duplicate what the platform or superpowers already handles well.
Before authoring, identify:
State the triggering situation in concrete terms. Prefer user intents, task conditions, and failure signals over broad domain labels.
Write the smallest set of instructions that reliably forces the move. Use direct commands, checks, and stop conditions. Avoid product-specific paths, local tool names, or host-only features unless they are documented as optional references.
Each real skill should include an eval.md alongside SKILL.md. The eval file records the empirical acceptance test from WORKFLOW.md: baseline behavior, with-skill behavior, what counts as improvement, and the pass bar.
For scaffold-only or template changes, direct mechanical validation is enough.
Optional files live next to SKILL.md when they remove noise from the main skill:
eval.md for the empirical acceptance test.scripts/ for executable helpers the model may run.reference.md for longer material loaded only when needed.Do not add empty support folders. Add them only when the skill actually needs them.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.