Provides runtime security enforcement and threat hunting for AI-driven code changes and system operations.
The Clawdstrike Security Review skill acts as an automated security auditor for Claude Code, ensuring that autonomous actions comply with organizational safety policies. It monitors for sensitive file access, dangerous shell commands, secret leaks, and unauthorized network egress by integrating with the Clawdstrike engine. By proactively checking actions against pre-defined guards like ForbiddenPath and SecretLeak, it prevents accidental exposure of credentials and unauthorized system modifications, making it an essential tool for managing autonomous AI fleets in sensitive production or development environments.
Key Features
01Proactive risk assessment of CI/CD and infrastructure configuration changes
02Filesystem protection via allowlists and forbidden path enforcement
03273 GitHub stars
04Automated secret scanning to prevent credential leaks in code and configuration
05Egress control to monitor and block unauthorized network requests
06Real-time shell command validation against customizable safety policies
Use Cases
01Scanning codebases for accidentally committed secrets, API keys, or private certificates
02Auditing shell commands executed by autonomous agents to prevent system compromise
03Reviewing and securing infrastructure-as-code changes before deployment
