Get recent user interactions (clicks, form submissions, input changes)
AI agents call get_user_interactions to retrieve information from Claude Imagine without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries and returns data about user interactions without creating, modifying, deleting, or executing any operations. It is a passive read operation with no ability to affect system state or trigger external actions. The low severity reflects minimal risk even if accessed by an untrusted agent, as it only exposes UI interaction history.
From the tool's definition Tool name 'get_user_interactions' and description explicitly states it retrieves recent user interactions (clicks, form submissions, input changes) with no modification or side effects indicated.
Attacks that exploit this kind of access
Get recent user interactions (clicks, form submissions, input changes). It is categorised as a Read tool in the Claude Imagine MCP Server, which means it retrieves data without modifying state.
Register the Claude Imagine MCP server in PolicyLayer and add a rule for get_user_interactions: 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 Claude Imagine. Nothing to install.
get_user_interactions 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_interactions 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_interactions. 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_interactions is provided by the Claude Imagine MCP server (t3rm1nu55/claudeimagine). 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 →