Perform deep research on a topic. Searches multiple sources, extracts citations, and synthesizes a comprehensive answer. Requires LLM sampling capability.
AI agents call search_deep to retrieve information from Mcp Open Webresearch without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool performs web searching and content extraction across multiple sources to synthesize information. It reads and aggregates data from the web without creating, modifying, or deleting any data. The 'medium' severity reflects the recursive/deep nature of the search which could access a wide range of web content and requires LLM sampling capability, but fundamentally remains a read operation.
From the tool's definition Searches multiple sources, extracts citations, and synthesizes a comprehensive answer
Attacks that exploit this kind of access
Perform deep research on a topic. Searches multiple sources, extracts citations, and synthesizes a comprehensive answer. Requires LLM sampling capability. It is categorised as a Read tool in the Mcp Open Webresearch MCP Server, which means it retrieves data without modifying state.
Register the Mcp Open Webresearch MCP server in PolicyLayer and add a rule for search_deep: 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 Open Webresearch. Nothing to install.
search_deep 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_deep 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_deep. 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_deep is provided by the Mcp Open Webresearch MCP server (rinaldowouterson/mcp-open-webresearch). 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 →