AI agents call diagnose_disagreement to retrieve information from Warrant without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool performs analysis on argumentation data to identify the source or nature of a disagreement between two agents. This is a read/query operation — it retrieves or computes an explanation but does not modify data, execute code, delete anything, or involve financial transactions. Severity is low as misuse would at worst produce an incorrect diagnosis.
From the tool's definition Diagnose WHY two agents disagree — this is an analytical/diagnostic operation that examines the nature of a disagreement
Attacks that exploit this kind of access
Diagnose WHY two agents disagree. It is categorised as a Read tool in the Warrant MCP Server, which means it retrieves data without modifying state.
Register the Warrant MCP server in PolicyLayer and add a rule for diagnose_disagreement: 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 Warrant. Nothing to install.
diagnose_disagreement 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 diagnose_disagreement 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 diagnose_disagreement. 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.
diagnose_disagreement is provided by the Warrant MCP server (jayden-chmod/warrant-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 →