AI agents call discover_field to retrieve information from Scholar without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool discovers and maps academic papers within a research field. This is purely informational retrieval with no side effects: it queries metadata about existing papers rather than creating, modifying, deleting, or executing operations. The blast radius of misuse is minimal—an agent might waste tokens or retrieve irrelevant papers, but cannot harm data or systems.
From the tool's definition Tool performs 'find surveys, foundational papers, and recent advances' — retrieval and discovery operations with no modification, deletion, or execution of external code.
Attacks that exploit this kind of access
Map a research field: find surveys, foundational papers, and recent advances. It is categorised as a Read tool in the Scholar MCP Server, which means it retrieves data without modifying state.
Register the Scholar MCP server in PolicyLayer and add a rule for discover_field: 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 Scholar. Nothing to install.
discover_field 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 discover_field 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 discover_field. 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.
discover_field is provided by the Scholar MCP server (liyux3/scholar-mcp). 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 →