codebase-memory-mcp
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a standard set of instructions for using Model Context Protocol (MCP) tools to analyze codebases. It promotes verification of tool outputs against local source files.
- [COMMAND_EXECUTION]: The skill mentions operations such as repository indexing (index_repository) and project deletion (delete_project), but explicitly instructs the agent not to execute these without user approval.
- [EXTERNAL_DOWNLOADS]: The instructions contain a specific safety rule preventing the agent from installing the codebase-memory-mcp server or other third-party skills during the workflow.
Audit Metadata