security-scan

Scanning Claude Code configurations involves checking CLAUDE.md, settings.json, mcp.json, hooks, and agents/*.md files for weaknesses. This audit identifies injection risks and misconfigurations, providing a comprehensive report to ensure new projects start with a hardened configuration baseline.

Yes, you need AgentShield installed to perform security scanning. It provides the core scanning engine required to audit your Claude Code configurations, detect vulnerabilities, and generate remediation guidance for common misconfigurations.

The best way to harden a new Claude Code project baseline is to run an AgentShield-based scan on your configuration files. This surfaces vulnerabilities and misconfigurations early, offering auto-fix and remediation guidance to secure your setup from the start.

Yes, you can get auto-fix and remediation guidance for common misconfigurations. The security scan evaluates your configuration files and provides actionable recommendations to correct identified security weaknesses and injection risks.

You should run a security scan when starting new Claude Code projects, updating existing configurations, or performing ongoing security hygiene checks. This ensures continuous protection against vulnerabilities and maintains a hardened configuration baseline.