# Generated by API Evangelist (build-phrasing.py). Our phrasing, not observed demand.
overlay: 1.0.0
info:
  title: API Evangelist conversational phrasing for ElevenLabs Conversational AI Agents API
  version: 1.0.0
extends: openapi/elevenlabs-agents-api-openapi.yml
actions:
- target: $.info
  update:
    x-apievangelist-phrasing:
      method: generated
      generated: '2026-09-26'
      generator: build-phrasing.py
      label: Generated by API Evangelist
      operations: 6
- target: $.paths['/v1/convai/agents'].get
  update:
    x-apievangelist-phrasing:
      intent: List my conversational AI agents
      effect: read
      questions:
      - Which voice agents have I built on my ElevenLabs account?
      - Can I page through my agents if I have a lot of them?
      instructions:
      - text: List all my conversational agents.
      - text: Show the next {page_size} agents after cursor {cursor}.
        slots:
          page_size: query.page_size
          cursor: query.cursor
      method: generated
      generated: '2026-09-26'
- target: $.paths['/v1/convai/agents'].post
  update:
    x-apievangelist-phrasing:
      intent: Create a conversational AI agent
      effect: write
      questions:
      - How do I create a voice agent with its own voice, model and tools?
      - What configuration does a new conversational agent need?
      instructions:
      - text: Create an agent named {name} with conversation config {conversation_config}.
        slots:
          name: requestBody.name
          conversation_config: requestBody.conversation_config
      - text: Set up a new conversational agent using {conversation_config}.
        slots:
          conversation_config: requestBody.conversation_config
      method: generated
      generated: '2026-09-26'
- target: $.paths['/v1/convai/agents/{agent_id}'].get
  update:
    x-apievangelist-phrasing:
      intent: Get an agent's configuration
      effect: read
      questions:
      - What voice, prompt and tools is one of my agents configured with?
      - Can I pull the full settings of a single agent?
      instructions:
      - text: Get the configuration of agent {agent_id}.
        slots:
          agent_id: path.agent_id
      - text: Show me how agent {agent_id} is set up.
        slots:
          agent_id: path.agent_id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/v1/convai/agents/{agent_id}'].delete
  update:
    x-apievangelist-phrasing:
      intent: Delete a conversational agent
      effect: destructive
      questions:
      - How do I permanently delete a voice agent?
      - Can a deleted agent's configuration be restored?
      instructions:
      - text: Delete agent {agent_id}.
        slots:
          agent_id: path.agent_id
      - text: Permanently remove agent {agent_id} and its configuration.
        slots:
          agent_id: path.agent_id
      method: generated
      generated: '2026-09-26'
- target: $.paths['/v1/convai/agents/{agent_id}'].patch
  update:
    x-apievangelist-phrasing:
      intent: Update an agent's configuration
      effect: write
      questions:
      - Can I change an agent's prompt or voice without resending the whole config?
      - How do I rename an existing agent?
      instructions:
      - text: Rename agent {agent_id} to {name}.
        slots:
          agent_id: path.agent_id
          name: requestBody.name
      - text: Update agent {agent_id} with conversation config {conversation_config}.
        slots:
          agent_id: path.agent_id
          conversation_config: requestBody.conversation_config
      method: generated
      generated: '2026-09-26'
- target: $.paths['/v1/convai/agents/{agent_id}/simulate'].post
  update:
    x-apievangelist-phrasing:
      intent: Simulate a conversation to test an agent
      effect: write
      questions:
      - Can I test my agent against a simulated caller before going live?
      - How many turns can a simulated agent conversation run?
      instructions:
      - text: Simulate a conversation with agent {agent_id}.
        slots:
          agent_id: path.agent_id
      - text: Run a simulated chat with agent {agent_id} as user {simulated_user_config} for up to {max_turns} turns.
        slots:
          agent_id: path.agent_id
          simulated_user_config: requestBody.simulated_user_config
          max_turns: requestBody.max_turns
      method: generated
      generated: '2026-09-26'