Public Scan Reports

# Clean Documentation Skill

## Source

- Input: `fixtures/benchmark/01-safe-documentation-skill`
- Source identity: `SKILL.md` SHA-256 `9456104ea9b33ff96d159de56350e361105561ae4a5c71127dd04252942aef2e`
- Scanner version: `0.1.1`

## Command

```bash
skillgate scan fixtures/benchmark/01-safe-documentation-skill --format json
```

## Capability Inventory

No capabilities were detected.

## Findings Summary

- Findings: `0`
- High or critical findings: `0`
- Scanned files: `1`

## Interpretation

This is expected behavior for a minimal documentation-only skill. It provides a
baseline example of a scan result that does not introduce shell, network, secret,
filesystem-write, MCP, prompt-override, or obfuscation review items.

## Suggested Policy Direction

No new capability approval is needed for this fixture. A repository adopting
SkillGate could start with nonblocking `skillgate scan .` and later add
`skillgate check . --policy skillgate.yaml` once expected capabilities are
known.

## Limitations

This report only describes the committed fixture content. It does not prove that
other documentation skills are safe, that future edits remain safe, or that a
runtime host cannot be misconfigured.

## What SkillGate Cannot Conclude

SkillGate does not execute the skill, call an LLM, prove intent, or prove the
absence of every possible risky instruction. It reports deterministic static
capability evidence for review.