Audit Claude Code configurations for security vulnerabilities using AgentShield.
npx skills add https://github.com/ignidus/iscagent --skill security-scan-ignidus
Please help me install this Agent Skill. Skill: security-scan Source: https://github.com/ignidus/iscagent/tree/main/skills/security-scan Command: npx skills add https://github.com/ignidus/iscagent --skill security-scan-ignidus
Audits Claude Code configurations to identify security vulnerabilities, misconfigurations, and risky patterns across project files using AgentShield.
Run the AgentShield scanner against your repository's Claude Code configuration to detect and remediate security issues.
High-intent search queries and answers about installing and using this skill.
Scanning MCP configurations for overly permissive permissions identifies risky access controls in your settings.json and mcp.json files. AgentShield validates these configs to prevent unauthorized tool access and potential data exfiltration.
Checking CLAUDE.md for insecure prompts involves scanning the file for risky auto-run instructions and prompt injection vulnerabilities. AgentShield inspects agent definitions and configuration files to detect and remediate these security issues.
Yes, AgentShield works with MCP configs and hooks directories by inspecting settings.json, mcp.json, hooks/, and agents/*.md files. It validates these configurations for potential data exfiltration and prompt injection risks.
The best way to scan repository configurations for prompt injection is running a dedicated security audit tool like AgentShield. It inspects agent definitions and hook scripts to identify potential injection vectors and data exfiltration risks.
Yes, you can export security audit findings from Claude Code configs in JSON, Markdown, HTML, and terminal formats. AgentShield generates these reports after scanning your repository for configuration vulnerabilities.
cescrafliScan .claude directory configurations for security vulnerabilities and misconfigurations.
TruCol270Automate codebase audits and apply tested fixes with atomic commits.
mkavalichAudit MCP servers for vulnerabilities, compliance, and operational risks.
cdalsoniiiDesigns and executes controlled failure experiments to validate system resilience via fault injection scenarios.
