仅供缺少原生 WebSearch/WebFetch 的第三方模型做上下文友好的资料概览；官方 Claude 应使用内置工具。
AI agents call research_brief to retrieve information from CC Web MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool retrieves and presents information in a summarized format, which is a Read operation. It has no capability to modify, delete, or execute actions. The blast radius of misuse is minimal—an AI agent would only retrieve potentially inaccurate or irrelevant information. Severity is low because the worst case is reading incorrect data, not data loss or execution of harmful commands.
From the tool's definition Tool is described as providing 'context-friendly information overview' (资料概览) for third-party models. The sibling tools are 'fetch_url', 'health_check', and 'web_search', all read-only operations.
Documented attack patterns abuse exactly the kind of access research_brief gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and CC Web MCP, and nothing reaches the server without passing your rules. This is the rule we recommend for research_brief:
{
"version": "1",
"default": "deny",
"tools": {
"research_brief": {}
}
}research_brief 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.
仅供缺少原生 WebSearch/WebFetch 的第三方模型做上下文友好的资料概览；官方 Claude 应使用内置工具。. It is categorised as a Read tool in the CC Web MCP MCP Server, which means it retrieves data without modifying state.
Register the CC Web MCP server in PolicyLayer and add a rule for research_brief: 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 CC Web MCP. Nothing to install.
research_brief 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 research_brief 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 research_brief. 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.
research_brief is provided by the CC Web MCP server (jcdizzy/cc-web-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from CC Web MCP, 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.
4 CC Web MCP tools catalogued and risk-classified — across an index of 43,000+ MCP servers.