Metacognitive operations: view the current thinking state, get strategy suggestions, switch strategies, and receive guidance on improving reasoning. The system automatically detects stuck states and recommends actions.
AI agents call metacog to retrieve information from Deep Thinker without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Metacog is introspective and diagnostic: it observes internal reasoning states, suggests improvements, and detects stuck conditions, but does not execute new reasoning paths, modify the thought graph, or trigger external operations. The guidance and suggestions are informational outputs, not actions. No data is created, modified, deleted, or externally acted upon.
From the tool's definition Tool performs 'view the current thinking state, get strategy suggestions' and 'receive guidance' — all read-only operations that retrieve information without modifying the reasoning graph or producing side effects.
Documented attack patterns abuse exactly the kind of access metacog gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Deep Thinker, and nothing reaches the server without passing your rules. This is the rule we recommend for metacog:
{
"version": "1",
"default": "deny",
"tools": {
"metacog": {}
}
}metacog 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.
Metacognitive operations: view the current thinking state, get strategy suggestions, switch strategies, and receive guidance on improving reasoning. The system automatically detects stuck states and recommends actions. It is categorised as a Read tool in the Deep Thinker MCP Server, which means it retrieves data without modifying state.
Register the Deep Thinker MCP server in PolicyLayer and add a rule for metacog: 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 Deep Thinker. Nothing to install.
metacog 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 metacog 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 metacog. 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.
metacog is provided by the Deep Thinker MCP server (nachosystems/deep-thinker). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Deep Thinker, 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.
17 Deep Thinker tools catalogued and risk-classified — across an index of 43,000+ MCP servers.