AI agents call kg_extract_judgments to retrieve information from Predicate without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
kg_extract_judgments retrieves and reports existing knowledge graph state (judgments about entities, schema slices). It has no side effects on the knowledge graph itself—it does not assert, modify, delete, or execute external operations. The instruction to 'distill this session' is guidance to the host model, not an action performed by the tool.
From the tool's definition Tool description indicates it 'Return[s]' judgments and schema information—a retrieval operation with no modification, creation, deletion, or external execution.
Attacks that exploit this kind of access
Return the j: schema slice, current judgments about touched entities, and a brief instructing you (the host model) to distill this session\. It is categorised as a Read tool in the Predicate MCP Server, which means it retrieves data without modifying state.
Register the Predicate MCP server in PolicyLayer and add a rule for kg_extract_judgments: 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 Predicate. Nothing to install.
kg_extract_judgments 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 kg_extract_judgments 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 kg_extract_judgments. 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.
kg_extract_judgments is provided by the Predicate MCP server (nordicagents/predicate). 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 →