Semantic search over everything that has appeared on the user
AI agents call search_screen_memory to retrieve information from Screen Memory without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and queries screen memory data via semantic search. It performs no write, delete, execute, or financial operations. The blast radius of misuse is limited to exposure of screen content the user has already seen; an AI cannot alter data or trigger external actions through search alone.
From the tool's definition Tool description states 'Semantic search over everything that has appeared on the user' — a retrieval operation with no modification capability.
Risk signalsBulk/mass operation — affects multiple targets
Attacks that exploit this kind of access
Semantic search over everything that has appeared on the user. It is categorised as a Read tool in the Screen Memory MCP Server, which means it retrieves data without modifying state.
Register the Screen Memory MCP server in PolicyLayer and add a rule for search_screen_memory: 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 Screen Memory. Nothing to install.
search_screen_memory 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 search_screen_memory 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 search_screen_memory. 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.
search_screen_memory is provided by the Screen Memory MCP server (yonglim2392/screen-memory-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 →