SECURITY RESOURCES / SKILL INSPECTION

SkillScan

AI agent skill security

151 rules15 categoriesv1.1.1 / cloud CLOUD API

SCAN A SKILL FILE

STATIC ANALYSIS

FILENAME CONTENT

0 BUTF-8

Select filesSKILL.md · scripts · configs512 KB / file · 20 files / batch

INSPECTION RESULTS

Awaiting inspection

Ready for inspection

No scan results

REGEX + RECURSIVE DECODING—

HOW SKILLSCAN WORKS

DETERMINISTIC / NO LLM

01UploadSkill file or scripts
02DiscoverSupported inputs
03DecodeBase64, hex, URL
04AnalyzeMatch security rules
05DeduplicateMerge findings
06ScoreCalculate file risk
07ReportReview and export

CLI

Local scanning. Offline by default.

node dist/cli.js ./path/to/skillCLI documentation →

Cloud API

Cloudflare-hosted static analysis.

POST /scanAPI documentation →

MCP

Agent-native security inspection.

node dist/cli.js setupMCP documentation →

NOTE / 001

No findings is not a guarantee of safety. Review skills before granting tools, credentials, or filesystem access. Partial scans are not complete assessments.