AI agents call search_grants to retrieve information from Grant without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and queries grant information from multiple external research funding databases. It has no side effects on data—it does not create, modify, delete, or execute any operations. The sibling tools (get_grant_detail, list_upcoming_deadlines) further confirm this is a read-only information retrieval service.
From the tool's definition Tool performs search and retrieval across grant databases ('search research grants across NSF, ERC, KRF/NRF'). Name and description indicate query/retrieval with no modification, creation, deletion, or execution of external operations.
Attacks that exploit this kind of access
Search research grants across NSF (US), ERC (EU), KRF/NRF (Korea). It is categorised as a Read tool in the Grant MCP Server, which means it retrieves data without modifying state.
Register the Grant MCP server in PolicyLayer and add a rule for search_grants: 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 Grant. Nothing to install.
search_grants 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_grants 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_grants. 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_grants is provided by the Grant MCP server (vertical-mcp/grant-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 →