Agent Security Hardening
Audits and secures AI agent workspaces by scanning for leaked secrets, verifying file permissions, and monitoring for unauthorized access.
The Agent Security skill provides a comprehensive suite of security audits designed to protect your Claude Code workspace from common vulnerabilities. It automatically scans for leaked credentials and API keys, ensures sensitive configuration files have restricted permissions, monitors the age of security tokens for rotation compliance, and prevents sensitive data from being accidentally committed to version control. This tool is essential for developers working with LLM agents that have access to private environments, ensuring that both credentials and workspace integrity remain compromise-free.
主な機能
01Network connectivity verification for authorized API endpoints
02Git exposure checks to prevent sensitive files from entering version control
03API key rotation tracking to monitor the age of security credentials
043 GitHub stars
05File permission auditing to identify world-readable credential files
06Automated secret scanning for leaked API keys, passwords, and tokens
ユースケース
01Automating daily security heartbeats to maintain workspace integrity
02Performing a pre-deployment security audit on an autonomous agent workspace
03Hardening local development environments against accidental credential exposure