# LOLSkills - Agent Skill Scanner by MagicSword

LOLSkills inspects agent skill bundles before you install them. NOVA and YARA detections, correlated evidence, and optional AI intent assessment.

## Product & Documentation

- Product: https://lolskills.io
- Documentation: https://lolskills.io/docs
- API: https://lolskills.io/docs/api
- Methodology: https://lolskills.io/methodology
- Install Protection: https://lolskills.io/docs/install-protection
- Discovery Archive: https://lolskills.io/discover
- Privacy: https://lolskills.io/privacy
- Credits: https://lolskills.io/credits

## Technical Details

The scanner uses NOVA Framework and YARA to inspect agent skills at exact GitHub commits or uploaded bundles. Reports include:
- Static analysis findings
- Correlated behavioral evidence
- Optional AI intent assessment (opt-in)
- Coverage reporting

Reports are unlisted bearer links. Submitted scans are separate from public discovery.

## Key Features

- No API key required for public service
- Scans pinned to immutable commits or file hashes
- Public discovery every 15 minutes
- Optional artifact inspection (ClamAV, LIEF, capa)
- Transparent coverage and incomplete status reporting

## Limits

This is static analysis. It may miss:
- Indirect behavior across files
- Obfuscated payloads
- Runtime-only actions
- Fetched dependencies

A completed scan describes one version under configured checks. It is not a guarantee of safety.

## Community

Built by MagicSword using NOVA Framework, YARA, YARA Forge, ClamAV, LIEF, and capa.
Community Slack: https://www.magicsword.io/slack-community

## License

Apache 2.0 (application repository)
Scanner images and MagicSword trademarks are not covered by an additional grant.