effect-mcp-server
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional, providing templates and patterns for building MCP servers. It uses well-known, legitimate libraries from the Effect ecosystem.
- [DATA_EXPOSURE]: While the skill demonstrates how to expose local files as resources (e.g., 'file:///README.md') and interact with databases, these are standard examples of MCP server functionality. There are no patterns suggesting unauthorized data harvesting or exfiltration to external domains.
- [REMOTE_CODE_EXECUTION]: No remote code execution patterns, such as piped shell commands or dynamic script downloads from untrusted sources, were detected.
- [PROMPT_INJECTION]: The instructions do not attempt to override the AI's safety guidelines or system prompts. The use of 'Critical' and 'Important' headers is purely for technical emphasis in a developer-centric context.
Audit Metadata