AI agents call audit_risk to retrieve information from Maple without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a read-only monitoring and analysis tool. It retrieves trace data, performs analysis (detection of shell/network/credentials anomalies), and returns insights in dashboard form. No side effects, modifications, code execution, or irreversible changes occur. The severity is low because misuse (e.g., scanning traces for sensitive data) poses limited direct harm—it remains a passive observer.
From the tool's definition The tool 'audits' and 'returns' a risk dashboard—it analyzes and reports on existing trace data without modifying, deleting, or executing anything.
Attacks that exploit this kind of access
Audit a trace for anomalies and risky behavior (shell/network/credentials) and return a risk dashboard. It is categorised as a Read tool in the Maple MCP Server, which means it retrieves data without modifying state.
Register the Maple MCP server in PolicyLayer and add a rule for audit_risk: 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 Maple. Nothing to install.
audit_risk 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 audit_risk 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 audit_risk. 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.
audit_risk is provided by the Maple MCP server (omar2001ramadan/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 →