skills.rest
adomot
adomotCOMMUNITY
Explore all skills by this author →
GitHub

security-scan

Audit Claude Code configurations for security vulnerabilities and injection risks.

Updated Mar 28, 2026
Search Tags:#security#vulnerability#injection#auditing#claude-code#agentshield
One-click install
npx skills add https://github.com/adomot/claude-settings --skill security-scan-adomot
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: security-scan
Source: https://github.com/adomot/claude-settings/tree/main/skills/aegis/security-scan
Command: npx skills add https://github.com/adomot/claude-settings --skill security-scan-adomot

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

AgentShieldを使用してClaude Code設定（.claude/ディレクトリ）のセキュリティ脆弱性、設定ミス、インジェクションリスクをスキャン。CLAUDE.md、settings.json、MCPサーバー、フック、エージェント定義をチェック。

Core Features & Use Cases

  • AgentShieldを用いたセキュリティ監査を自動化し、機密情報の露出・不適切な権限・プロンプトインジェクションのリスクを検出します。
  • CLAUDE.md、settings.json、mcp.json、hooks、agentsディレクトリの構成要素を横断的に検査し、設定ミスを特定します。
  • Use Case: 新規 Claude Code プロジェクトのオンボーディング時や設定変更後のセキュリティ衛生チェックに適用します。

Quick Start

Run a scan using AgentShield in your Claude Code project root to perform an initial security assessment.

Frequently Asked Questions about security-scan

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I audit Claude Code configurations for security vulnerabilities?▼

You can audit Claude Code configurations by scanning the .claude directory to detect security vulnerabilities, misconfigurations, and injection risks. This includes checking CLAUDE.md, settings.json, mcp.json, hooks, and agent definitions for exposed secrets and risky permissions.

What is AgentShield and how does it secure Claude Code settings?▼

AgentShield is a security auditing tool that scans the .claude directory to identify secrets, risky permissions, MCP server risks, and prompt-injection vectors. It automates security hygiene checks across CLAUDE.md, settings.json, mcp.json, hooks, and agent definitions.

When should I run a security scan on my .claude directory?▼

You should run a security scan during new Claude Code project onboarding, after making configuration changes, and as periodic hygiene checks. This ensures CLAUDE.md, settings.json, and MCP server settings remain free of injection risks and misconfigurations.

Can I detect prompt-injection vectors in MCP server configurations?▼

Yes, you can detect prompt-injection vectors by scanning mcp.json and other MCP server configurations in the .claude directory. The security audit identifies risky permissions and injection risks across your Claude Code setup using AgentShield.

What's the best way to check for secrets and risky permissions in Claude Code hooks?▼

The best way to check for secrets and risky permissions in Claude Code hooks is to perform an automated security audit on the .claude directory. This scans hooks, agent definitions, and settings.json to detect exposed credentials and misconfigurations.

Related Skills

Explore More Software Engineering Skills→
bitwardenbitwarden
129

reviewing-claude-config

Review Claude Code configuration files for security risks and quality issues.

Official
Basic
MomoDavilukeMomoDaviluke
2

security-scan

Scan Claude Code configurations for vulnerabilities and misconfigurations using AgentShield.

Community
Advanced
kk20300113-pngkk20300113-png

security-scan

Scan Claude Code configuration files for security vulnerabilities and misconfigurations.

Community
Intermediate
kivo360kivo360
1

moai-domain-security

Mitigate security vulnerabilities in CI/CD pipelines using OWASP Top 10 and SAST/DAST workflows.

Community
Advanced