AI agents call vapi_get_assistant to retrieve information from Mcp Vapi without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves configuration details about an existing assistant. It is a read-only query operation with no side effects, no code execution capability, and no data modification. The low severity reflects minimal risk even if an AI agent misuses it, as it only exposes existing configuration data without capability to alter systems or trigger external operations.
From the tool's definition Tool description states 'Get full details of a specific Vapi assistant by ID' — a pure retrieval operation with 'Get' indicating query-only semantics. No modification, deletion, execution, or financial impact.
Attacks that exploit this kind of access
Get full details of a specific Vapi assistant by ID, including its model, voice, and prompt config. It is categorised as a Read tool in the Mcp Vapi MCP Server, which means it retrieves data without modifying state.
Register the Mcp Vapi MCP server in PolicyLayer and add a rule for vapi_get_assistant: 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 Mcp Vapi. Nothing to install.
vapi_get_assistant 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 vapi_get_assistant 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 vapi_get_assistant. 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.
vapi_get_assistant is provided by the Mcp Vapi MCP server (adhirajhangal/mcp-vapi). 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 →