Structured web search via Novada (Google). Returns titles, URLs, and descriptions \u2014 no HTML parsing needed. Best for finding pages and factual queries. For reading a specific URL, use agentproxy_fetch instead.
AI agents call agentproxy_search to retrieve information from Novada Proxy without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
While the agentproxy_search tool itself is a read-only operation (retrieving search results), the severity is elevated to 'medium' due to the proxy infrastructure that enables bypassing anti-bot protections and geo-restrictions.
From the tool's definition Tool performs 'Structured web search' returning 'titles, URLs, and descriptions' — retrieves data without modification.
Attacks that exploit this kind of access
Structured web search via Novada (Google). Returns titles, URLs, and descriptions \u2014 no HTML parsing needed. Best for finding pages and factual queries. For reading a specific URL, use agentproxy_fetch instead. It is categorised as a Read tool in the Novada Proxy MCP Server, which means it retrieves data without modifying state.
Register the Novada Proxy MCP server in PolicyLayer and add a rule for agentproxy_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 Novada Proxy. Nothing to install.
agentproxy_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 agentproxy_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 agentproxy_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.
agentproxy_search is provided by the Novada Proxy MCP server (novadalabs/novada-proxy). 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 →