security-scan

To scan Claude Code configurations for security vulnerabilities, you can use this Skill to automate the detection of prompt-injection vectors, hardcoded secrets, and risky automation patterns across your project files. It evaluates CLAUDE.md, settings.json, MCP configurations, hooks, and agent definitions to surface high-priority security issues.

Security scanning for Claude Code settings identifies overly permissive permissions, hardcoded secrets, and risky automation patterns. It validates your settings.json, MCP configurations, hooks, and agent definitions to prevent potential security incidents caused by misconfigurations.

Yes, you need AgentShield installed to execute this security scan on Claude Code files. The Skill relies on AgentShield to perform the underlying analysis of your CLAUDE.md, settings.json, and MCP configurations to detect vulnerabilities and report findings.

Yes, you can integrate this Claude Code security scanning into CI/CD pipelines and CLI workflows. The Skill supports standard outputs, allowing you to automate security checks during pre-commit validations, new project onboarding, and periodic security hygiene assessments.

Automated security auditing scans CLAUDE.md files to identify sensitive data exposure and prompt injection vectors. By analyzing the configuration content, it proactively reports vulnerabilities and misconfigurations to help teams prevent security incidents before deployment.