AI agents invoke click_screen to trigger actions in Computer Control 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.
Clicking on screen coordinates is an Execute-category action because it triggers external UI operations whose effects are entirely argument-dependent. A misused click could submit forms, confirm deletions, authorize transactions, or trigger any number of irreversible actions depending on the current screen state. The blast radius is high because the agent controls a real computer UI.
From the tool's definition "Click at the specified screen coordinates" — triggers a mouse click action on the screen, which is a browser/UI interaction that can activate buttons, links, and controls with effects depending on what is clicked.
Documented attack patterns abuse exactly the kind of access click_screen gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Computer Control MCP, and nothing reaches the server without passing your rules. This is the rule we recommend for click_screen:
{
"version": "1",
"default": "deny",
"tools": {
"click_screen": {
"limits": [
{
"counter": "click_screen_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}click_screen 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.
Free to start. No card required.
Click at the specified screen coordinates. It is categorised as a Execute tool in the Computer Control MCP MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Computer Control MCP server in PolicyLayer and add a rule for click_screen: 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 Computer Control MCP. Nothing to install.
click_screen 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 click_screen 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 click_screen. 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.
click_screen is provided by the Computer Control MCP server (ab498/computer-control-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 15 Computer Control MCP tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
15 Computer Control MCP tools catalogued and risk-classified — across an index of 42,500+ MCP servers.