AI agents call hound_search to retrieve information from MCP-Hound without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs read-only operations on indexed code repositories. It retrieves and queries data using regex patterns but cannot create, modify, delete, or execute code. The worst-case misuse would be unauthorized information disclosure about code structure and content, which is a low-severity read risk. The blast radius is limited to exposing information that already exists in indexed repositories.
From the tool's definition Tool description explicitly states it "Search[es] code across all indexed repositories" and "Returns matching files and lines". The verb "search" combined with "returns" indicates data retrieval with no modification capability.
Attacks that exploit this kind of access
Search code across all indexed repositories using regex patterns. Returns matching files and lines with deep links. It is categorised as a Read tool in the MCP-Hound MCP Server, which means it retrieves data without modifying state.
Register the MCP-Hound MCP server in PolicyLayer and add a rule for hound_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-Hound. Nothing to install.
hound_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 hound_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 hound_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.
hound_search is provided by the MCP-Hound MCP server (jmagly/mcp-hound). 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 →