Skip to main content
Glama
elevenlabs-agents-mcp-server

elevenlabs-agents-mcp-server

elevenlabs-agents-mcp-server

MCP server that lets Claude place outbound telephone calls via ElevenLabs Agents (Conversational AI) over a SIP trunk (e.g. Vonage) and report the outcome — status, success evaluation, summary, extracted data, and full transcript.

Tools

Tool

Purpose

elevenlabs_list_agents

Discover agent IDs

elevenlabs_list_phone_numbers

Discover SIP trunk phone number IDs

elevenlabs_make_outbound_call

Place a call (returns conversation_id immediately)

elevenlabs_wait_for_call_completion

Poll until the call ends, then return the outcome report

elevenlabs_get_call_report

Compact outcome report (status, call_successful, summary, data collection, cost, termination reason)

elevenlabs_get_call_transcript

Turn-by-turn transcript with pagination

elevenlabs_list_calls

Recent calls with filters (agent, outcome, time window)

Typical flow: list_agents → list_phone_numbers → make_outbound_call → wait_for_call_completion → (optionally) get_call_transcript.

Related MCP server: leximo-ai-call-assistant-mcp-server

Setup

npm install
npm run build

Claude Desktop / Claude Code config

{
  "mcpServers": {
    "elevenlabs-agents": {
      "command": "node",
      "args": ["/absolute/path/to/elevenlabs-agents-mcp-server/dist/index.js"],
      "env": {
        "ELEVENLABS_API_KEY": "sk_..."
      }
    }
  }
}

Claude Code: claude mcp add elevenlabs-agents -e ELEVENLABS_API_KEY=sk_... -- node /path/to/dist/index.js

Vonage SIP trunk prerequisites (one-time, in dashboards — not via this server)

  1. Vonage: create a SIP trunk / programmable SIP domain. For outbound-from-ElevenLabs you need a termination endpoint (hostname) that accepts SIP INVITEs from ElevenLabs, with digest auth credentials. Allow signaling from sip.rtc.elevenlabs.io (or your region's static SIP endpoint).

  2. ElevenLabs → Phone Numbers → Import from SIP trunk: enter your Vonage number (E.164), outbound address (Vonage termination hostname), transport (TLS recommended), and SIP credentials.

  3. Assign or note the agent; the imported number's phone_number_id is what elevenlabs_make_outbound_call uses as agent_phone_number_id.

  4. Analysis fields (call_successful, summary, data collection) require evaluation criteria / data collection configured on the agent.

Environment variables

  • ELEVENLABS_API_KEY (required)

  • ELEVENLABS_API_BASE_URL (optional, default https://api.elevenlabs.io; use residency endpoints if applicable)

Notes

  • make_outbound_call is asynchronous — the API returns as soon as the call is initiated. Post-call analysis is only available once status is done (there is a short processing window after hang-up).

  • wait_for_call_completion caps at 900 s per invocation and tells you if it timed out so you can call it again.

  • Dynamic variables passed to make_outbound_call are available in the agent's prompt as {{variable_name}}.

Available Tools

7 tools
elevenlabs_get_call_reportGet Call Outcome ReportA
Read-onlyIdempotent

Get a compact outcome report for a call/conversation: status, success evaluation, summary, duration, termination reason, extracted data-collection results, and evaluation criteria results. Does NOT include the transcript (use elevenlabs_get_call_transcript).

Note: analysis fields (call_successful, summary, data collection) are only populated after the call ends and post-call processing completes (status 'done'). If status is 'initiated'/'in-progress'/'processing', poll again or use elevenlabs_wait_for_call_completion.

Args:

  • conversation_id (string)

Returns JSON: { conversation_id, status, call_successful, summary, duration_secs, termination_reason, error, cost_credits, evaluation_criteria_results, data_collection_results, transcript_turns, has_audio }

Inputs: conversation_id (required).

elevenlabs_get_call_transcriptGet Call TranscriptA
Read-onlyIdempotent

Get the turn-by-turn transcript of a call.

Args:

  • conversation_id (string)

  • offset / limit: pagination over turns (default 0 / 200)

  • include_tool_calls (boolean): default false

Returns JSON: { conversation_id, status, total_turns, has_more, turns: [{ index, role, time_in_call_secs, message }] }

Inputs: conversation_id (required), limit, offset, include_tool_calls.

elevenlabs_list_agentsList ElevenLabs AgentsA
Read-onlyIdempotent

List conversational AI agents in the ElevenLabs workspace.

Use this first to discover the agent_id needed by elevenlabs_make_outbound_call.

Args:

  • search (string, optional): filter by agent name

  • page_size (number): 1-100, default 30

  • cursor (string, optional): pagination cursor

Returns JSON: { agents: [{ agent_id, name, created_at_unix_secs }], has_more, next_cursor }

Inputs: cursor, search, page_size.

elevenlabs_list_callsList Recent CallsA
Read-onlyIdempotent

List recent conversations/calls, newest first, with optional filters.

Args:

  • agent_id (string, optional)

  • call_successful ('success' | 'failure' | 'unknown', optional)

  • call_start_after_unix / call_start_before_unix (number, optional)

  • page_size (1-100, default 30), cursor (string, optional)

Returns JSON: { calls: [{ conversation_id, agent_id, agent_name, status, call_successful, start_time_unix_secs, call_duration_secs, direction }], has_more, next_cursor }

Inputs: cursor, agent_id, page_size, call_successful, +2 more.

elevenlabs_list_phone_numbersList ElevenLabs Phone NumbersA
Read-onlyIdempotent

List phone numbers imported into ElevenLabs (including SIP trunk numbers, e.g. a Vonage SIP trunk DID).

Use this to discover the agent_phone_number_id required by elevenlabs_make_outbound_call. SIP trunk numbers have provider 'sip_trunk'.

Args: none

Returns JSON: { phone_numbers: [{ phone_number_id, phone_number, label, provider, assigned_agent }] }

Inputs: no named parameters.

elevenlabs_make_outbound_callPlace Outbound Call (SIP Trunk)A

Place an outbound telephone call via an ElevenLabs agent over a SIP trunk (e.g. Vonage).

This starts the call and returns immediately — it does NOT wait for the call to finish. To get the outcome, follow up with elevenlabs_wait_for_call_completion using the returned conversation_id.

Args:

  • agent_id (string): agent that will speak on the call

  • agent_phone_number_id (string): SIP trunk number ID to originate from

  • to_number (string): destination in E.164 format, e.g. +61355551234

  • dynamic_variables (object, optional): variables for the agent's prompt

  • first_message_override (string, optional)

  • prompt_override (string, optional; agent must allow overrides)

Returns JSON: { success, message, conversation_id, sip_call_id } Save conversation_id — it is needed to retrieve the outcome and transcript.

Error handling: 422 usually means a bad to_number format or invalid IDs.

Inputs: agent_id (required), to_number (required), agent_phone_number_id (required), prompt_override, +2 more.

elevenlabs_wait_for_call_completionWait for Call Completion and Report OutcomeA
Read-only

Block until a call reaches a terminal status ('done' or 'failed') or the timeout elapses, then return the outcome report (same shape as elevenlabs_get_call_report).

This is the recommended follow-up after elevenlabs_make_outbound_call: place the call, then call this to get the result in one step.

Args:

  • conversation_id (string)

  • timeout_secs (number): default 300

  • poll_interval_secs (number): default 5

Returns JSON: { timed_out, waited_secs, ...outcome report } If timed_out is true, the call is still in progress — call this tool again to keep waiting.

Inputs: conversation_id (required), timeout_secs, poll_interval_secs.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.4/5.0

    Scored across 7 tools

    Disambiguation5/5

    Each tool has a unique, well-defined purpose: call reports, transcripts, agent listing, call listing, phone number listing, making calls, and waiting for completion. No overlap in functionality.

    Naming Consistency5/5

    All tools follow the 'elevenlabs_verb_noun' pattern with snake_case, e.g., elevenlabs_get_call_report, elevenlabs_list_agents. Perfectly consistent.

    Tool Count5/5

    Seven tools is a tight set covering listing, call initiation, and post-call retrieval. No redundant tools, and scope is well-defined for an outbound call management server.

    Completeness4/5

    The workflow is complete: discover agents/numbers, place call, wait, get results. Minor gaps like no tool to create or update agents/numbers, but those may be out of scope.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers