AI agents call check_leak to retrieve information from Mullvad without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves or queries the status of potential VPN leaks (DNS, IP, WebRTC, etc.) to inform the user, but does not modify any settings, execute commands, or delete data. It is a read-only diagnostic check. The empty description slightly lowers confidence, but the context from the server description and tool name are sufficiently clear that this is a passive information-gathering operation.
From the tool's definition Tool name is 'check_leak' and the server description indicates it's part of VPN management suite for checking 'for leaks' — a diagnostic/monitoring operation with no side effects.
Attacks that exploit this kind of access
check_leak. It is categorised as a Read tool in the Mullvad MCP Server, which means it retrieves data without modifying state.
Register the Mullvad MCP server in PolicyLayer and add a rule for check_leak: 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 Mullvad. Nothing to install.
check_leak 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 check_leak 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 check_leak. 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.
check_leak is provided by the Mullvad MCP server (oresam-xyz/mullvad-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 →