mcp-python-conventions-taudelta
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
- [SAFE]: The skill provides coding conventions for MCP servers, emphasizing protocol correctness and security by recommending logging to stderr to avoid corrupting the JSON-RPC stream and using explicit error handling to prevent sensitive data leakage.
- [EXTERNAL_DOWNLOADS]: The skill references documentation and repositories from trusted and well-known organizations, including Anthropic, the Model Context Protocol project, and Prefect.
- [DATA_EXFILTRATION]: The skill includes instructions to prevent accidental data exposure by advocating for the use of ToolError to sanitize exception messages before they reach the client, ensuring internal system details are not leaked during error conditions.
Audit Metadata