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, andsudo. 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
.envfiles, credentials, and hardcoded API keys. - [EXTERNAL_DOWNLOADS]: Mentions the use of
npx cc-safefor auditing projects and@company/db-mcpas 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