Skill Security Auditor — Security agent for Claude Code
Audits an untrusted skill folder (its SKILL.md and any scripts) for instructions that would read secrets, send data out, run destructive commands or pull in remote instructions.
How to install Skill Security Auditor
Installs to ~/.claude/agents/naman10parikh-skillsmith-skill-security-auditor.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/naman10parikh/skillsmith/HEAD/.claude/agents/skill-security-auditor.md -o ~/.claude/agents/naman10parikh-skillsmith-skill-security-auditor.md Restart Claude Code, or start a new session, for it to be picked up.
What Skill Security Auditor does
name: skill-security-auditor description: Audits an untrusted skill folder (its SKILL.md and any scripts) for instructions that would read secrets, send data out, run destructive commands or pull in remote instructions. Use before installing a skill from outside your own repositories, or when skillsmith install or a marketplace brings in a new skill. allowed-tools: ["Read", "Glob", "Grep", "Bash"]
You audit skills for Skillsmith before they reach a skills folder. A skill is a set of ins
Alternatives in Security
- AI Code Auditor — PYA specialist spawned by the /pya orchestrator 240 ★
- Bats Test Validator — Validates BATS test comprehensiveness and integrity for bash scripts 129 ★
- Knowledge Base Loader — Phase KB0 intake agent that converts staged, untrusted application documentation into a cited, security-orient 125 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Skill Vetter
Read-only static security auditor for untrusted Claude Code skills, subagents, commands, and plugins. Use BEFO
Sigil PR Audit
Use this agent to audit untrusted Sigil pull requests behavior-first before reading implementation details.
PR Security Review
Use this agent to perform a security analysis of a pull request BEFORE checking out or running any code locall
N2i Sec Scanner
Runs the n2i-sec-audit deterministic scan (scripts/scan.py) and returns a short result so the long scan log st
Kenaz
Audits the security of Claude Code plugins, MCP servers, and AI agents before installation. Analyzes executabl
Repo Audit
Audits exactly ONE repository per invocation (code: repo-audit) and writes a graded Security / Gaps / Improvem
Related Skills
My Security Review
Run a personal agent-tooling security review - secrets, shell/hook command execution, untrusted input, plugin/
Claude Config Auditor
Non-destructive linter and cost auditor for Claude Code's .claude/ and CLAUDE.md — measures token cost, audits
Skill Scorecard
A 30-second safety scorecard for AI agent skills: run it on any skill folder or GitHub repo before installing.