AI agents invoke exit_playmode to trigger actions in Unity MCP. 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 an editor state transition (stopping Play Mode), which is an external operation affecting the Unity Editor environment. It's not a simple read, nor does it delete/modify persistent data, but it does trigger a real action in an external application. Any unsaved Play Mode changes would be lost, making it potentially disruptive, though it's described as a 'safe way to stop gameplay testing'.
From the tool's definition 'Exit Unity Play Mode and return to Edit Mode' — triggers an external operation in the Unity Editor changing its runtime state
Attacks that exploit this kind of access
Exit Unity Play Mode and return to Edit Mode. Safe way to stop gameplay testing. It is categorised as a Execute tool in the Unity MCP MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Unity MCP server in PolicyLayer and add a rule for exit_playmode: 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 Unity MCP. Nothing to install.
exit_playmode 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 exit_playmode 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 exit_playmode. 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.
exit_playmode is provided by the Unity MCP server (muammar-yacoob/unity-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 →