AI agents invoke cmux_new_pane to trigger actions in Cmux. 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 triggers external operations (terminal and browser execution) whose effects depend on subsequent user/agent input. While the tool itself only creates the pane container, it enables Execute-category activities (shell command execution, browser automation) without built-in safeguards.
From the tool's definition Tool opens new panes (terminal or browser) by splitting. Terminal panes enable execution of arbitrary shell commands; browser panes enable navigation to and interaction with websites.
Attacks that exploit this kind of access
Open a new pane (terminal or browser) by splitting in the given direction. It is categorised as a Execute tool in the Cmux MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Cmux MCP server in PolicyLayer and add a rule for cmux_new_pane: 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 Cmux. Nothing to install.
cmux_new_pane 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 cmux_new_pane 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 cmux_new_pane. 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.
cmux_new_pane is provided by the Cmux MCP server (puchkoff/cmux-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 →