AI agents call swarm_vector_search to retrieve information from Mcp Swarm without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a read-only operation that searches and retrieves information. It performs vector similarity queries against indexed code or documents, returning matching results. There are no side effects, state changes, or destructive operations. The blast radius of misuse is minimal—an agent could retrieve unintended documents, but cannot execute code, modify data, or cause other harm.
From the tool's definition Tool name 'swarm_vector_search' and description '使用向量相似度搜索代码或文档' (Perform vector similarity search on code or documents) indicate a search/query operation.
Attacks that exploit this kind of access
使用向量相似度搜索代码或文档。. It is categorised as a Read tool in the Mcp Swarm MCP Server, which means it retrieves data without modifying state.
Register the Mcp Swarm MCP server in PolicyLayer and add a rule for swarm_vector_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 Mcp Swarm. Nothing to install.
swarm_vector_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 swarm_vector_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 swarm_vector_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.
swarm_vector_search is provided by the Mcp Swarm MCP server (moselu/mcp-swarm). 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 →