ui_undo · UI Bridge MCP: Risk & Policy

WHAT IT DOES

What ui_undo does on UI Bridge MCP

AI agents invoke ui_undo to trigger actions in UI Bridge 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.

WHY IT NEEDS A POLICY

Why ui_undo needs a policy

This tool executes an operation (undo) on the runner's UI state. While undo is reversible by nature, it still qualifies as Execute rather than Write because it triggers an external operation whose effects depend on what was previously done (i.e., what state the UI is in). The tool has moderate blast radius if misused in a loop or during critical operations, but is mitigated by being reversible.

From the tool's definition Tool description states it 'Execute[s] undo in the runner', and it is listed among interaction tools on a UI control server that supports interactive control mode. Undo is a reversible state-change operation that executes an action within the runner's own UI.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

ui_undo Rate-limited

ui_undo stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "ui_undo": {
      "limits": [
        {
          "counter": "ui_undo_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

RATE-LIMIT THIS TOOL →

Instant setup, no code required.

PolicyLayer is an MCP gateway: every tool call is checked against your rule before it reaches UI Bridge MCP, so you can run agents against it safely.

FAQ

Questions about ui_undo

What does the ui_undo tool do? +

Execute undo in the runner. It is categorised as a Execute tool in the UI Bridge MCP MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on ui_undo? +

Register the UI Bridge MCP server in PolicyLayer and add a rule for ui_undo: 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 UI Bridge MCP. Nothing to install.

What risk level is ui_undo? +

ui_undo is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.

Can I rate-limit ui_undo? +

Yes. Add a rate_limit block to the ui_undo 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.

How do I block ui_undo completely? +

Set action: deny in the PolicyLayer policy for ui_undo. 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.

What MCP server provides ui_undo? +

ui_undo is provided by the UI Bridge MCP server (qontinui/ui-bridge-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.