Search persistent memories (hybrid BM25 + vector semantic retrieval). Use when you need to recall previous context, patterns, decisions, or bug fix records.
AI agents call memory_search to retrieve information from Claude Persistent Memory without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and queries stored memory data without creating, modifying, deleting, or executing anything. It is clearly a Read category tool. Severity is medium rather than low because the memories may contain sensitive context, decisions, or bug fix records that could inform an agent's behavior in harmful ways if poisoned or exfiltrated, but the tool itself has no direct destructive potential.
From the tool's definition Tool performs search/retrieval operations ('Search persistent memories', 'recall previous context') with no modification or execution capability. The hybrid BM25 + vector semantic search mechanism is a query-only operation.
Documented attack patterns abuse exactly the kind of access memory_search gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Claude Persistent Memory, and nothing reaches the server without passing your rules. This is the rule we recommend for memory_search:
{
"version": "1",
"default": "deny",
"tools": {
"memory_search": {}
}
}memory_search is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Search persistent memories (hybrid BM25 + vector semantic retrieval). Use when you need to recall previous context, patterns, decisions, or bug fix records. It is categorised as a Read tool in the Claude Persistent Memory MCP Server, which means it retrieves data without modifying state.
Register the Claude Persistent 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 Claude Persistent 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 Claude Persistent Memory MCP server (mimi180306/claude-persistent-memory). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Claude Persistent Memory, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
4 Claude Persistent Memory tools catalogued and risk-classified — across an index of 43,000+ MCP servers.