Search the UI tree for elements whose label, value, or hint contains query text
AI agents call find_elements to retrieve information from App Screen without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a query/search tool that retrieves information from the UI accessibility tree without side effects. It matches the Read category pattern of retrieving or querying data. The blast radius is minimal—misuse results only in retrieving unintended UI information, not system changes, financial impact, or code execution.
From the tool's definition The tool 'find_elements' performs a search operation on the UI tree, retrieving elements that match query criteria. The description explicitly states it 'Search[es] the UI tree for elements' with no mention of modifications, deletions, or code execution.
Attacks that exploit this kind of access
Search the UI tree for elements whose label, value, or hint contains query text. It is categorised as a Read tool in the App Screen MCP Server, which means it retrieves data without modifying state.
Register the App Screen MCP server in PolicyLayer and add a rule for find_elements: 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 App Screen. Nothing to install.
find_elements 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 find_elements 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 find_elements. 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.
find_elements is provided by the App Screen MCP server (xmuweili/app-screen-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 →