skill-scanner 0.1.1

Security scanner for AI agent skills and instruction artifacts

skill-scanner

skill-scanner reviews AI skill and instruction artifacts for security risk using:

• OpenAI analysis
• VirusTotal analysis

Requirements

• Python 3.11+
• uv
• OpenAI and/or VirusTotal API key (at least one)

Install (from source)

uv sync --all-extras --group dev

Run with:

uv run skill-scanner --help

Alias:

uv run skillscan --help

What gets scanned

By default, discover and scan detect markdown-based skill/instruction artifacts (for example SKILL.md, AGENTS.md, CLAUDE.md, *.instructions.md, *.prompt.md, *.agent.md, .mdc).

Use --path to target a specific file or folder.

Quick start

# See targets
uv run skill-scanner discover --format json

# Verify key/model configuration
uv run skill-scanner doctor

# Run a combined scan (if both keys are configured)
uv run skill-scanner scan --format summary

Key configuration and analyzer selection

scan requires at least one analyzer enabled.

• If only OPENAI_API_KEY is available, AI runs and VT is disabled.
• If only VT_API_KEY is available, VT runs and AI is disabled.
• If both keys are available, VT findings are included and VT context is passed into AI analysis.
• You can disable either analyzer with --no-ai or --no-vt.

API key safety

Use 1Password secret references instead of plaintext secrets:

OPENAI_API_KEY=op://Developer/OpenAI/api_key
VT_API_KEY=op://Developer/VirusTotal/api_key

Run the scanner through 1Password CLI so references are resolved at runtime:

op run --env-file=.env -- uv run skill-scanner scan --format summary

Security best practice:

• Prefer a 1Password Service Account scoped to only the vault/items required for scanning (least privilege).

Reference:

• https://developer.1password.com/docs/cli/secret-references/

Output formats

scan --format supports:

• table (default)
• summary
• json
• sarif

You can write output to a file with --output <path>.

Useful commands

# List providers
uv run skill-scanner providers

# Scan one path only
uv run skill-scanner scan --path ./some/skill/folder --format summary

# List discovered targets without running analyzers
uv run skill-scanner scan --list-targets

# Scan only selected discovered targets (repeat --target)
uv run skill-scanner scan --target /absolute/path/to/SKILL.md --target /absolute/path/to/AGENTS.md --format summary

# Filter to medium+
uv run skill-scanner scan --min-severity medium --format summary

# Non-zero exit if high+ findings exist
uv run skill-scanner scan --fail-on high --format summary

--list-targets can be used without API keys because it only runs discovery and exits.

Exit behavior

• 0: scan completed and fail threshold not hit
• 1: --fail-on threshold matched
• 2: no analyzers enabled (for example missing keys combined with flags), or --target did not match any discovered target