grepai-mcp-claude
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates semantic search and call graph analysis over local codebases, creating a surface where malicious instructions embedded in the code could influence the agent's behavior.
- Ingestion points: Project codebase files indexed via the 'grepai watch' command.
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded content in search results.
- Capability inventory: The skill provides tools for semantic search (grepai_search), call graph tracing (grepai_trace_callers, grepai_trace_callees, grepai_trace_graph), and index status (grepai_index_status).
- Sanitization: No explicit sanitization or filtering of codebase content is described.
- [COMMAND_EXECUTION]: The documentation instructs users to execute CLI commands to register the MCP server.
- Evidence: Commands such as 'claude mcp add grepai -- grepai mcp-serve' and 'grepai mcp-serve' are suggested for setup and verification. These are standard operational procedures for the specified integration.
Audit Metadata