Search through indexed documentation using full-text search, you should try also query with synonyms.
AI agents call search_documentation to retrieve information from Whoosh RAG MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves information from a pre-built documentation index without creating, modifying, deleting, or executing anything. It is a pure query operation with no side effects. The sibling tools (build_documentation_index, update_documentation_index) handle data modification, but this tool only reads.
From the tool's definition Tool performs 'Search through indexed documentation using full-text search' with no mention of modification, deletion, or execution capabilities. The description explicitly indicates retrieval ('find', 'search') of existing indexed data.
Attacks that exploit this kind of access
Search through indexed documentation using full-text search, you should try also query with synonyms. It is categorised as a Read tool in the Whoosh RAG MCP MCP Server, which means it retrieves data without modifying state.
Register the Whoosh RAG MCP server in PolicyLayer and add a rule for search_documentation: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Whoosh RAG MCP. Nothing to install.
search_documentation is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the search_documentation rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for search_documentation. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
search_documentation is provided by the Whoosh RAG MCP server (jianlins/whoosh_rag_mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →