AI agents call poi_search to retrieve information from Attom without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
POI search is fundamentally a query operation that retrieves location-based data without modifying or deleting records. However, confidence is moderate (0.75) due to the empty tool description, which limits certainty. The severity is medium because bulk POI location data could potentially be used for reconnaissance or mapping sensitive locations, though the risk is lower than write/execute/destructive operations.
From the tool's definition Tool name 'poi_search' indicates a search operation for points of interest. The sibling tools on this server (all_events_detail, assessment_detail, boundary_detail, etc.) are all Read-category data retrieval operations.
Attacks that exploit this kind of access
poi_search. It is categorised as a Read tool in the Attom MCP Server, which means it retrieves data without modifying state.
Register the Attom MCP server in PolicyLayer and add a rule for poi_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 Attom. Nothing to install.
poi_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 poi_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 poi_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.
poi_search is provided by the Attom MCP server (nkbud/mcp-server-attom). 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 →