AI agents call search_citations to retrieve information from Eff without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries a bibliography database and returns results without creating, modifying, deleting, or executing anything. It is a straightforward read operation with minimal blast radius—worst case, an agent could retrieve irrelevant citation metadata, causing no harm.
From the tool's definition Tool description explicitly states it 'Search[es] BibTeX entries for a query string' and 'Returns minimal metadata.' The verb 'search' combined with 'returns' indicates data retrieval with no modification or side effects.
Attacks that exploit this kind of access
Search BibTeX entries for a query string in any field. Returns minimal metadata. It is categorised as a Read tool in the Eff MCP Server, which means it retrieves data without modifying state.
Register the Eff MCP server in PolicyLayer and add a rule for search_citations: 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 Eff. Nothing to install.
search_citations 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_citations 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_citations. 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_citations is provided by the Eff MCP server (vs3kulic/eff-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 →