必要に応じてWeb検索を実行し、出典付きの根拠で回答を返す
AI agents call answer to retrieve information from Openai Responses without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries the web and provides answers with citations, which is a read-only operation that retrieves and presents information. There are no side effects that create, modify, delete, or execute arbitrary code. The web search is scoped to answering questions, making it a safe Read category tool with low severity risk.
From the tool's definition Tool performs web search and returns answers with citations/sources. Description indicates "execute web search as needed" and "return answers with sources" — purely information retrieval with no data modification, deletion, or external state changes.
Attacks that exploit this kind of access
必要に応じてWeb検索を実行し、出典付きの根拠で回答を返す. It is categorised as a Read tool in the Openai Responses MCP Server, which means it retrieves data without modifying state.
Register the Openai Responses MCP server in PolicyLayer and add a rule for answer: 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 Openai Responses. Nothing to install.
answer 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 answer 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 answer. 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.
answer is provided by the Openai Responses MCP server (uchimanajet7/openai-responses-mcp). 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 →