Use the tool to think about something. It will not obtain new information or change the database, but just append the thought to the log. Use it when complex reasoning or some cache memory is needed.
AI agents call think to retrieve information from Think MCP Tool without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool explicitly states it does not obtain new information or change the database. It only records internal reasoning to a log, making it a benign read/note-taking operation with negligible blast radius.
From the tool's definition 'It will not obtain new information or change the database, but just append the thought to the log'
Documented attack patterns abuse exactly the kind of access think gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Think MCP Tool, and nothing reaches the server without passing your rules. This is the rule we recommend for think:
{
"version": "1",
"default": "deny",
"tools": {
"think": {}
}
}think 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.
Use the tool to think about something. It will not obtain new information or change the database, but just append the thought to the log. Use it when complex reasoning or some cache memory is needed. It is categorised as a Read tool in the Think MCP Tool MCP Server, which means it retrieves data without modifying state.
Register the Think MCP Tool MCP server in PolicyLayer and add a rule for think: 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 Think MCP Tool. Nothing to install.
think 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 think 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 think. 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.
think is provided by the Think MCP Tool MCP server (rai220/think-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 4 Think MCP Tool tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
4 Think MCP Tool tools catalogued and risk-classified — across an index of 42,500+ MCP servers.