Find entities semantically similar to a natural-language query, then optionally expand via
AI agents call graph_search to retrieve information from Graph-Memory without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves data from the graph based on semantic similarity without creating, modifying, deleting, or executing external operations. It is purely informational and has no side effects on the underlying knowledge graph or system state.
From the tool's definition Tool name 'graph_search' and description 'Find entities semantically similar to a natural-language query' indicate retrieval/query operations with no modification capability.
Attacks that exploit this kind of access
Find entities semantically similar to a natural-language query, then optionally expand via. It is categorised as a Read tool in the Graph-Memory MCP Server, which means it retrieves data without modifying state.
Register the Graph-Memory MCP server in PolicyLayer and add a rule for graph_search: 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 Graph-Memory. Nothing to install.
graph_search 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 graph_search 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 graph_search. 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.
graph_search is provided by the Graph-Memory MCP server (stevepridemore/graph-memory). 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 →