ML model asset discovery and size reporting.
AI agents call audit_ml_assets to retrieve information from Code-Oracle without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs asset discovery and generates reports on ML model sizes, which are passive information-gathering operations with no side effects. It retrieves and queries data without creating, modifying, deleting, or executing anything. The context of an architectural/security auditing server reinforces that this is an analytical tool for inventory and metrics collection.
From the tool's definition Tool name includes 'audit' and description states 'discovery and size reporting' — both indicate read-only data retrieval operations. No modification, deletion, or execution capabilities are described.
Attacks that exploit this kind of access
ML model asset discovery and size reporting. It is categorised as a Read tool in the Code-Oracle MCP Server, which means it retrieves data without modifying state.
Register the Code-Oracle MCP server in PolicyLayer and add a rule for audit_ml_assets: 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 Code-Oracle. Nothing to install.
audit_ml_assets 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_ml_assets 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_ml_assets. 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_ml_assets is provided by the Code-Oracle MCP server (sri-lohith-mulugu/oracle-server). 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 →