AI agents call search_tools to retrieve information from MCPFind 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 performs semantic search across tool schemas and descriptions. It has no side effects, does not execute code or modify data, and only returns information about other tools that are available. The blast radius is minimal since misuse would only return irrelevant search results.
From the tool's definition Tool name 'search_tools' and description 'Search available tools by natural language description' indicate a query/discovery operation that retrieves metadata about available tools without modifying or executing them.
Attacks that exploit this kind of access
Search available tools by natural language description. It is categorised as a Read tool in the MCPFind MCP Server, which means it retrieves data without modifying state.
Register the MCPFind MCP server in PolicyLayer and add a rule for search_tools: 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 MCPFind. Nothing to install.
search_tools 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 search_tools 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 search_tools. 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.
search_tools is provided by the MCPFind MCP server (jcgs2503/mcpfind). 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 →