memory_search
AI agents call memory_search to retrieve information from Eve Memory without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
A search tool retrieves data with no direct side effects. However, severity is elevated to medium because it queries a shared cross-agent memory system that may contain sensitive preferences, rules, and accumulated context from multiple agents. Misuse could enable information disclosure or manipulation of agent behavior through memory inspection.
From the tool's definition Tool name 'memory_search' combined with sibling tools like 'memory_get_preferences', 'memory_extract', and 'memory_ingest' indicates a shared memory retrieval function. The name 'search' suggests querying/retrieving data without modification.
Attacks that exploit this kind of access
memory_search. It is categorised as a Read tool in the Eve Memory MCP Server, which means it retrieves data without modifying state.
Register the Eve Memory MCP server in PolicyLayer and add a rule for memory_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 Eve Memory. Nothing to install.
memory_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 memory_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 memory_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.
memory_search is provided by the Eve Memory MCP server (sherifkozman/eve-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 →