Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.mrassistant.ai/llms.txt

Use this file to discover all available pages before exploring further.

An agent is the unit of configuration for a voice experience: instructions, model selection, voice, tools, and knowledge sources. You create and update agents through the HTTP API and manage them per tenant.

​
Lifecycle

  1. Create an agent with baseline instructions and provider defaults.
  2. Iterate using partial updates as you tune prompts and behavior.
  3. Attach knowledge when answers should be grounded in your documents or URLs.
  4. Enable tools when the agent must call your APIs or built-in actions.
  5. Run conversations or sessions against a stable agent id.

​
API surface

The generated API reference lists routes under /agents (including documents, knowledge, tools, conversations, and sessions). Use request and response schemas from OpenAPI when building clients.

​
Design tips

  • Write spoken instructions: short sentences, explicit confirmation steps, and clear escalation paths.
  • Keep tool contracts small and testable; validate arguments server-side.
  • Use multilingual settings when callers may switch languages mid-session.

Knowledge & tools

MCP for B2C