AI agents call get_user_decision to retrieve information from AFK Mode without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool sends a prompt to the mobile client and waits for a response, ultimately returning the user's decision. The primary effect is reading/retrieving a user's choice rather than modifying data, executing code, or causing destructive effects. However, it has a side effect of sending a notification, which is a minor Write-like action.
From the tool's definition 'Sends a decision request to the mobile client and blocks until the user responds or timeout expires' — it retrieves a user decision/response, acting as a query that waits for and returns user input.
Attacks that exploit this kind of access
Sends a decision request to the mobile client and blocks until the user responds or timeout expires. Only call when AFK mode is active. Use type. It is categorised as a Read tool in the AFK Mode MCP Server, which means it retrieves data without modifying state.
Register the AFK Mode MCP server in PolicyLayer and add a rule for get_user_decision: 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 AFK Mode. Nothing to install.
get_user_decision 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 get_user_decision 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 get_user_decision. 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.
get_user_decision is provided by the AFK Mode MCP server (paulbennet/afk-mode-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 →