🌟 RECOMMENDED: Intelligent query router with automatic strategy selection. Classifies query type (symbol lookup, concept explanation, how-to, etc.) and automatically chooses optimal retrieval method (code index vs documentation RAG). Returns grounded answers with full reasoning trace showing whi...
AI agents call smart_query to retrieve information from RAG Document Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
smart_query performs semantic search and question-answering over documents using vector embeddings. It routes queries intelligently but does not create, modify, delete, or execute code. The reasoning trace it returns is metadata about retrieval logic, not execution of arbitrary operations.
From the tool's definition Tool description states it "classifies query type" and "automatically chooses optimal retrieval method" with "returns grounded answers". The verb phrases indicate information retrieval only: no modification, deletion, or execution of external operations.
Attacks that exploit this kind of access
🌟 RECOMMENDED: Intelligent query router with automatic strategy selection. Classifies query type (symbol lookup, concept explanation, how-to, etc.) and automatically chooses optimal retrieval method (code index vs documentation RAG). Returns grounded answers with full reasoning trace showing which tools were called and why. Use this as your primary entry point instead of choosing between individual tools. It is categorised as a Read tool in the RAG Document Server MCP Server, which means it retrieves data without modifying state.
Register the RAG Document Server MCP server in PolicyLayer and add a rule for smart_query: 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 RAG Document Server. Nothing to install.
smart_query 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 smart_query 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 smart_query. 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.
smart_query is provided by the RAG Document Server MCP server (jaimeferj/mcp-rag-docs). 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 →