SkillScan by RakTikoo

Scan an AI agent skill or MCP server before you install it.

AI-agent skills and MCP servers run on your machine, with your agent's permissions — your files, your accounts, your keys. SkillScan reads a skill folder before you install it and gives you one plain-English verdict — BLOCK, CAUTION, or PASS — from deterministic rules.

By default no LLM is involved and nothing leaves your machine: the same folder always gets the same answer, and every finding points at the exact rule, file, and line. Scanned code is never executed — files are only read and parsed. An optional LLM review mode (--llm, off by default) can add findings on top; it can never remove or change the offline ones.

Other projects share the name SkillScan. This one is RakTikoo's, and the skillscan name on PyPI belongs to a different tool by another author — please use the GitHub routes below.

The verdicts

Exit codeVerdictMeaning
0PASSNo known-bad patterns were found in what was scanned.
1CAUTIONSomething worth a human read before installing.
2BLOCKAt least one block-severity pattern was found.
3—Usage / IO error: the scanner could not run (bad path, bad flags). Deliberately distinct from BLOCK.

The exit codes make SkillScan usable as a CI / install-time gate.

Install

SkillScan is pure Python 3 standard library, with zero dependencies (Python 3.10+). Install it as a command with pip, straight from GitHub:

pip install git+https://github.com/RakTikoo/skillscan
skillscan <path-to-skill-folder>

With pipx (isolated) or uv:

pipx install git+https://github.com/RakTikoo/skillscan
uv tool install git+https://github.com/RakTikoo/skillscan

From the release wheel (no git needed):

pip install https://github.com/RakTikoo/skillscan/releases/download/v0.10.0/skillscan-0.10.0-py3-none-any.whl

Checksums for the release files are in the SHA256SUMS file attached to the v0.10.0 release.

With Homebrew (macOS/Linux):

brew tap RakTikoo/tools
brew install RakTikoo/tools/skillscan

With Scoop (Windows):

scoop bucket add RakTikoo https://github.com/RakTikoo/scoop-bucket
scoop install skillscan

Or clone the repo and run it in place:

git clone https://github.com/RakTikoo/skillscan
cd skillscan
python3 -m skillscan <path-to-skill-folder>

Guides

What this is not

SkillScan is a linter, not a guarantee of safety.

Known limits, kept honest:

Full details, the rule table, and the complete known-limits list are in the GitHub repository.