Keyword-based web search returning a list of search results. Each result includes title, description, and URL. Best for quickly scanning multiple web pages or when you need to see diverse sources. Returns up to 20 search results as a list, not synthesized answers.
AI agents call brave_web_search to retrieve information from Claude TypeScript MCP Servers without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a read-only operation that queries a web search engine and returns information. It has no side effects, does not execute code, does not modify data, and does not commit financial transactions. The tool merely retrieves and presents search results to the user. Blast radius is minimal—misuse would only result in undesired web queries, not damage to systems or data.
From the tool's definition Tool description states it returns "a list of search results" with "title, description, and URL" and is "Best for quickly scanning multiple web pages". It explicitly retrieves data without modifying or executing actions.
Documented attack patterns abuse exactly the kind of access brave_web_search gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Claude TypeScript MCP Servers, and nothing reaches the server without passing your rules. This is the rule we recommend for brave_web_search:
{
"version": "1",
"default": "deny",
"tools": {
"brave_web_search": {}
}
}brave_web_search is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Keyword-based web search returning a list of search results. Each result includes title, description, and URL. Best for quickly scanning multiple web pages or when you need to see diverse sources. Returns up to 20 search results as a list, not synthesized answers. It is categorised as a Read tool in the Claude TypeScript MCP Servers MCP Server, which means it retrieves data without modifying state.
Register the Claude TypeScript MCP Servers MCP server in PolicyLayer and add a rule for brave_web_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 Claude TypeScript MCP Servers. Nothing to install.
brave_web_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 brave_web_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 brave_web_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.
brave_web_search is provided by the Claude TypeScript MCP Servers MCP server (ukkz/claude-ts-mcps). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Claude TypeScript MCP Servers, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
84 Claude TypeScript MCP Servers tools catalogued and risk-classified — across an index of 43,000+ MCP servers.