Stops a running tunnel or all local tunnels. This tool will: - Stop a specific tunnel identified by name (if provided) - Stop all local tunnels (if no name is provided) - Only affects tunnels running on the current machine - Will not affect tunnels running on other machines After stopping tunnels...
AI agents invoke stop_tunnel to trigger actions in Mcp Untun. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
This tool executes commands that terminate running processes/services on the local machine. While not destructive (tunnels can be restarted via start_tunnel), it is irreversible in the immediate term and disrupts active network operations. The dual mode (stop by name or all tunnels) means argument-dependent effects that could impact multiple services simultaneously if misused by an agent.
From the tool's definition Tool performs 'Stops a running tunnel or all local tunnels' and 'will stop a specific tunnel identified by name' or 'Stop all local tunnels' — these are operations that trigger external systems (tunnel processes) and whose effects depend on runtime arguments…
Attacks that exploit this kind of access
Stops a running tunnel or all local tunnels. This tool will: - Stop a specific tunnel identified by name (if provided) - Stop all local tunnels (if no name is provided) - Only affects tunnels running on the current machine - Will not affect tunnels running on other machines After stopping tunnels, you can use. It is categorised as a Execute tool in the Mcp Untun MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Mcp Untun MCP server in PolicyLayer and add a rule for stop_tunnel: 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 Mcp Untun. Nothing to install.
stop_tunnel is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the stop_tunnel 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 stop_tunnel. 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.
stop_tunnel is provided by the Mcp Untun MCP server (minte-app/untun-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 →