AI agents call session_search to retrieve information from Ensemble without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a straightforward query/search operation that retrieves data matching semantic criteria and returns results. No data is created, modified, deleted, or executed. The operation is read-only with no side effects. Low severity because search operations on session data have minimal blast radius even if results are misused by an agent.
From the tool's definition Tool description explicitly states 'Search sessions by semantic similarity. Returns top-K matches with scores.' The verb 'search' and the function of returning matches with scores indicates a retrieval operation with no modification or side effects.
Attacks that exploit this kind of access
Search sessions by semantic similarity. Returns top-K matches with scores. It is categorised as a Read tool in the Ensemble MCP Server, which means it retrieves data without modifying state.
Register the Ensemble MCP server in PolicyLayer and add a rule for session_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 Ensemble. Nothing to install.
session_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 session_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 session_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.
session_search is provided by the Ensemble MCP server (lynkbyte/ensemble). 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 →