AI agents call ask-one-question to retrieve information from AskMeMCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves information (human responses) without modifying, executing external operations, deleting data, or moving money. It is a query/read operation that enables human-in-the-loop decision-making by requesting clarification. The severity is low because misuse results only in unnecessary interruption of execution flow, not data loss, code execution, or other harmful side effects.
From the tool's definition Tool name 'ask-one-question' and description indicate it 'Get[s] free-form text responses' from human input.
Attacks that exploit this kind of access
Get free-form text responses when a situation is very unclear and cannot be solved with the other tools. Use ONLY when facing total uncertainty where you don\. It is categorised as a Read tool in the AskMeMCP MCP Server, which means it retrieves data without modifying state.
Register the AskMe MCP server in PolicyLayer and add a rule for ask-one-question: 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 AskMeMCP. Nothing to install.
ask-one-question 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 ask-one-question 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 ask-one-question. 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.
ask-one-question is provided by the AskMe MCP server (roman-vm/ask-me-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 →