From skill-linting
Review ~/.claude/settings.json for risky auto-allowed permissions. Use when auditing a Claude Code configuration for safety, or when asked to check that the agent cannot run code, delete files, reach the network, or read secrets without confirmation.
How this skill is triggered — by the user, by Claude, or both
Slash command
/skill-linting:allow-deny-safetyThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Check ~/.claude/settings.json. Make sure that the agent can not do any of the following automatically:
Check ~/.claude/settings.json. Make sure that the agent can not do any of the following automatically:
python or bash commands)cat /etc/passwd or ls ~)curl or requests library)apt install or pip install or brew install)AWS_SECRET_KEY or DATABASE_URL), or read or source the .env fileCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub ryanpeach/skills --plugin skill-linting