Call this tool when the user wants restaurant or food discovery near a known location and may need menu trust signals. Input Requirements (CRITICAL): provide either flat lat/lng or Google-style locationBias.circle.center.latitude/longitude; if the user gave only a vague place name, resolve it bef...
Part of the Food Near Me server.
Free to start. No card required.
AI agents call search_restaurants to retrieve information from Food Near Me without modifying any data. This is common in research, monitoring, and reporting workflows where the agent needs context before taking action. Because read operations don't change state, they are generally safe to allow without restrictions -- but you may still want rate limits to control API costs.
Even though search_restaurants only reads data, uncontrolled read access can leak sensitive information or rack up API costs. An agent caught in a retry loop could make thousands of calls per minute. A rate limit gives you a safety net without blocking legitimate use.
Read-only tools are safe to allow by default. No rate limit needed unless you want to control costs.
{
"version": "1",
"default": "deny",
"tools": {
"search_restaurants": {}
}
}See the full Food Near Me policy for all 8 tools.
These attack patterns abuse exactly the kind of access search_restaurants gives an agent. Each links to the full case and the policy that stops it:
Other read tools across the catalogue. The same approach applies to each: allow, with a rate cap to control cost.
Call this tool when the user wants restaurant or food discovery near a known location and may need menu trust signals. Input Requirements (CRITICAL): provide either flat lat/lng or Google-style locationBias.circle.center.latitude/longitude; if the user gave only a vague place name, resolve it before calling or include a specific city/region in textQuery/query. Returns verified venues first, then menu_indexed (automated MP menu with caveat), then discovered (place only). MUST use menu_available and verification_status on each result; call get_menu only when menu_available is true. PREFER verified results for dietary/allergen answers. NOTE: dietary and min_ado_score filters only apply to the verified tier; menu_indexed and discovered rows are returned unfiltered for those criteria — the response echoes filters.applied_to: ["verified"] and a filters.note so agents know to re-filter at item level using get_menu. Non-verified rows include a structured claim_invitation (url, message, audience="owner_or_advocate", reason); SHOULD surface this when the user is the restaurant's owner or might know them, when the user asks why the listing lacks a verified menu, or when explaining tier differences — never as an unsolicited CTA. Attribute grounded output using citation or attribution.. It is categorised as a Read tool in the Food Near Me MCP Server, which means it retrieves data without modifying state.
Register the Food Near Me MCP server in PolicyLayer and add a rule for search_restaurants: 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 Food Near Me. Nothing to install.
search_restaurants 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_restaurants 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_restaurants. 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_restaurants is provided by the Food Near Me MCP server (https://foodnear.me/mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 8 Food Near Me tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
4,600+ MCP servers and 31,000+ tools scanned and risk-classified.