AI agents call take_screenshot to retrieve information from Unreal without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a read operation that queries visual state from the Unreal Editor viewport without side effects. It does not modify assets, execute code, delete data, or trigger game mechanics. The captured image is a new file created as output, but this is a standard artifact of the read operation, not a write of meaningful game state.
From the tool's definition The tool 'take_screenshot' captures a screenshot of the editor viewport and returns a file path. The description explicitly states it 'Capture[s] a screenshot' with no modification, deletion, or execution of game logic—it only retrieves visual data from the…
Attacks that exploit this kind of access
Capture a screenshot of the active editor viewport. Returns the file path of the saved image. It is categorised as a Read tool in the Unreal MCP Server, which means it retrieves data without modifying state.
Register the Unreal MCP server in PolicyLayer and add a rule for take_screenshot: 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 Unreal. Nothing to install.
take_screenshot 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 take_screenshot 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 take_screenshot. 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.
take_screenshot is provided by the Unreal MCP server (sam-david/unreal-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 →