Delete a Mistral agent by ID
AI agents call delete_agent to permanently remove resources in Mistral Agent Manager — typically in cleanup and lifecycle workflows. It does its job in a single call, and there is no undo.
This tool permanently deletes agents without the ability to restore them. While the blast radius is limited to the specific agent being deleted (not system-wide), the irreversible nature of deletion and the potential impact on dependent workflows or automations classifies this as Destructive rather than Write. Severity is high because deletion of production agents could disrupt critical AI operations.
From the tool's definition Tool name is 'delete_agent' and description states 'Delete a Mistral agent by ID'. The delete operation is irreversible and removes agent configurations from the Mistral system.
Attacks that exploit this kind of access
Delete a Mistral agent by ID. It is categorised as a Destructive tool in the Mistral Agent Manager MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the Mistral Agent Manager MCP server in PolicyLayer and add a rule for delete_agent: 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 Mistral Agent Manager. Nothing to install.
delete_agent is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the delete_agent 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 delete_agent. 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.
delete_agent is provided by the Mistral Agent Manager MCP server (uzbyr/mcp-server). 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 →