agent-tool-builder
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive guidance on tool design, JSON schema definitions, and the Model Context Protocol (MCP) using established developer practices.
- [SAFE]: Code examples for Python and TypeScript use well-known, legitimate SDKs from Anthropic, OpenAI, and Vercel.
- [SAFE]: The content includes explicit security recommendations for developers, such as warning against SQL injection and emphasizing the importance of input validation and timeouts for external calls.
- [SAFE]: No patterns associated with obfuscation, data exfiltration, persistence, or unauthorized command execution were found.
Audit Metadata