agent-configuration

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents several dangerous command patterns, such as rm -rf /, chmod 777, and sudo. These are used exclusively within security 'Hooks' examples to demonstrate how to configure an agent to block or warn against destructive actions.
  • [DATA_EXFILTRATION]: Provides explicit 'DO NOT' policies and security guidelines to prevent the exposure of sensitive data, such as .env files, credentials, and hardcoded API keys.
  • [EXTERNAL_DOWNLOADS]: Mentions the use of npx cc-safe for auditing projects and @company/db-mcp as a placeholder for team MCP servers. These are presented in the context of security auditing and official team configuration, following standard development practices for the Claude Code platform.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 05:51 AM
Security Audit — agent-trust-hub — agent-configuration