Submit a targeted ROI OCR command to the running observer daemon. The daemon captures the window region, runs OCR, and stores the result. Non-blocking — returns a command ID you can poll with a second call.
AI agents call observer_ocr_roi to retrieve information from ScreenHand without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
OCR is a read operation that extracts text from visual content without modifying, executing code, deleting data, or creating financial transactions. However, severity is elevated to 'medium' rather than 'low' because OCR on arbitrary window regions could expose sensitive information (passwords, tokens, PII, financial data) depending on what applications are running.
From the tool's definition Tool performs OCR (Optical Character Recognition) on a window region and retrieves/stores the result for querying. The description states it 'runs OCR' and returns data that 'you can poll,' indicating information retrieval with no side effects on the system.
Documented attack patterns abuse exactly the kind of access observer_ocr_roi gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and ScreenHand, and nothing reaches the server without passing your rules. This is the rule we recommend for observer_ocr_roi:
{
"version": "1",
"default": "deny",
"tools": {
"observer_ocr_roi": {}
}
}observer_ocr_roi is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Submit a targeted ROI OCR command to the running observer daemon. The daemon captures the window region, runs OCR, and stores the result. Non-blocking — returns a command ID you can poll with a second call. It is categorised as a Read tool in the ScreenHand MCP Server, which means it retrieves data without modifying state.
Register the ScreenHand MCP server in PolicyLayer and add a rule for observer_ocr_roi: 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 ScreenHand. Nothing to install.
observer_ocr_roi 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 observer_ocr_roi 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 observer_ocr_roi. 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.
observer_ocr_roi is provided by the ScreenHand MCP server (manushi4/screenhand). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from ScreenHand, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
89 ScreenHand tools catalogued and risk-classified — across an index of 43,000+ MCP servers.