ElevenLabs Agents MCP server

Read from the package source without running it. The installed server may list more.

• add_agent_tool
• add_knowledge_base_document
• cancel_batch_callwrite action

  Cancel a queued or scheduled batch-call job before it dials more recipients. WHEN TO USE: - Stop a scheduled job that should no longer run - Halt a queued batch after noticing a mistake in the audience or timing EXAMPLE: {"batch_id": "batch_123"} RELATED TOOLS: - get_batch_call: confirm the current job status before cancelling - list_batch_calls: find the correct batch_id - submit_batch_call: crea

• check_order_statuswrite action

  Look up an order by ID.

• configure_elevenlabs_agents_api_key

  Host-managed setup only. The user adds the ElevenLabs Agents API key in Settings → Connectors in the app. Do not ask for or accept the key in chat. WHEN TO USE: - Only when the host supplies the key during connector setup EXAMPLE: - Host-managed setup supplies the key directly after the user saves it in Settings → Connectors RELATED TOOLS: - list_agents: inspect agents after configuring RETURNS: o

• create_agentwrite action
• delete_agentwrite action

  Permanently delete one ElevenLabs agent. WHEN TO USE: - Remove a temporary or test agent such as a rebel-live-test-* artifact - Clean up a duplicate or abandoned draft after confirming it is no longer needed EXAMPLE: {"agent_id": "agent_123"} RELATED TOOLS: - duplicate_agent: safer way to branch before destructive edits - get_agent: confirm the exact agent before deletion RETURNS: ok confirmation.

• delete_knowledge_base_documentwrite action

  Delete one knowledge-base document or folder from ElevenLabs. WHEN TO USE: - Remove a temporary rebel-live-test-* document after validation - Clean up a stale or incorrect document before re-adding it EXAMPLE: {"documentation_id": "doc_123", "force": true} RELATED TOOLS: - get_knowledge_base_doc: confirm the exact document before deleting it - add_knowledge_base_document: re-create the document af

• delete_phone_numberwrite action

  Permanently delete one imported phone number from your ElevenLabs workspace. WHEN TO USE: - Remove a number that was imported by mistake - Decommission a line before releasing it at the telephony provider EXAMPLE: {"phone_number_id": "pn_123"} RELATED TOOLS: - get_phone_number: confirm the exact number before deleting it - list_phone_numbers: verify the number is gone afterwards RETURNS: ok confir

• duplicate_agent

  Duplicate one ElevenLabs agent, optionally with a new name. WHEN TO USE: - Make a safe copy before editing a production agent - Fork an existing agent as the starting point for a variant EXAMPLE: {"agent_id": "agent_123", "name": "Support triage v2"} RELATED TOOLS: - get_agent: inspect the duplicated config in full - update_agent: make targeted edits on the duplicate - delete_agent: remove the dup

• get_agent

  Get full configuration for one ElevenLabs agent, including prompts and nested conversation settings. WHEN TO USE: - Inspect the system prompt or first message before changing anything - Confirm the exact agent configuration behind recent conversations EXAMPLE: {"agent_id": "agent_123"} RELATED TOOLS: - list_agents: discover valid agent IDs - list_conversations: inspect recent conversations for thi

• get_batch_call

  Get one batch-call job, including per-recipient statuses and dynamic variables. WHEN TO USE: - Check which recipients succeeded, failed, or are still queued - Confirm the stored scheduled time before deciding to cancel or retry EXAMPLE: {"batch_id": "batch_123"} RELATED TOOLS: - list_batch_calls: discover valid batch IDs - cancel_batch_call: stop this job if it is still queued or scheduled - retry

• get_conversation

  Get full details for one conversation, including transcript, caller-provided content, and analysis. WHEN TO USE: - Review a single conversation in detail after listing recent conversations - Inspect transcript text before escalating or making follow-up changes EXAMPLE: {"conversation_id": "conv_123"} RELATED TOOLS: - list_conversations: discover valid conversation IDs - get_conversation_audio: dow

• get_conversation_audio

  Download the audio recording for one conversation to a temporary local file. WHEN TO USE: - After get_conversation when you need the original audio - To preserve evidence from a conversation before sharing it elsewhere EXAMPLE: {"conversation_id": "conv_123"} RELATED TOOLS: - list_conversations: discover conversation IDs - get_conversation: inspect transcript and analysis alongside the audio RETUR

• get_knowledge_base_doc

  Get one knowledge-base document: metadata from GET /knowledge-base/{id}, plus body text from GET /knowledge-base/{id}/content when available. WHEN TO USE: - Inspect a specific document before later add/delete work - Review metadata (type, URL sync flags) and the text agents can retrieve EXAMPLE: {"documentation_id": "doc_123"} RELATED TOOLS: - list_knowledge_base_docs: discover valid documentation

• get_knowledge_base_rag_index_status

  Get the RAG index status for one knowledge-base document, so you can tell when uploaded content is retrievable by agents. WHEN TO USE: - After add_knowledge_base_document, to confirm indexing has finished before testing retrieval - To diagnose why an agent is not using a document's content EXAMPLE: {"documentation_id": "doc_123"} RELATED TOOLS: - add_knowledge_base_document: upload the document fi

• get_phone_number

  Get one phone number, including its label and assigned agent information. WHEN TO USE: - Confirm which agent is assigned to a phone number - Inspect telephony setup before later write-side phone updates EXAMPLE: {"phone_number_id": "pn_123"} RELATED TOOLS: - list_phone_numbers: discover valid phone_number_id values - get_agent: inspect the agent assigned to this phone number RETURNS: phone_number.

• import_phone_number
• list_agent_tools

  List workspace tools available to your ElevenLabs agents (webhook, client, and system tools). WHEN TO USE: - Discover tool IDs before wiring a tool into an agent's configuration - Inventory existing webhook/client tools before adding a new one EXAMPLE: {"page_size": 10} EXAMPLE: {"search": "calendar"} RELATED TOOLS: - add_agent_tool: create a webhook or client tool when none exists yet - update_ag

• list_agents

  List voice agents in your ElevenLabs Conversational AI workspace. WHEN TO USE: - Discover available agent IDs before inspecting one in detail - Inventory check before reviewing prompts or recent conversations EXAMPLE: {"page_size": 10} RELATED TOOLS: - get_agent: inspect a returned agent_id in full - list_conversations: review recent calls for a returned agent_id RETURNS: agents, count, next_curso

• list_batch_calls

  List batch-call jobs in your ElevenLabs workspace. WHEN TO USE: - Review recent scheduled or completed batch jobs - Find a batch_id before fetching one job in detail EXAMPLE: {"limit": 10} RELATED TOOLS: - get_batch_call: inspect one batch job in detail - cancel_batch_call: stop a queued or scheduled job - retry_batch_call: re-run a failed or partial batch RETURNS: batch_calls, count, next_cursor.

• list_conversations

  List conversations in your ElevenLabs Conversational AI workspace. WHEN TO USE: - Find a recent conversation ID before pulling the full transcript - Review calls for one agent over a time range EXAMPLE: {"agent_id": "agent_123", "page_size": 5} RELATED TOOLS: - get_conversation: fetch the full transcript and analysis for one conversation - get_conversation_audio: download the recording for a retur

• list_knowledge_base_docs

  List knowledge-base documents in your ElevenLabs Conversational AI workspace. WHEN TO USE: - Discover document IDs before fetching one document in full - Inventory the current knowledge base attached to your voice-agent workspace EXAMPLE: {"page_size": 10} RELATED TOOLS: - get_knowledge_base_doc: fetch one returned documentation_id in detail - list_agents: inspect which agents might rely on these

• list_phone_numbers

  List phone numbers configured in your ElevenLabs Conversational AI workspace. WHEN TO USE: - Find available phone_number_id values before inspecting one number in detail - Review labels and agent assignments on the telephony side EXAMPLE: {"page_size": 10} RELATED TOOLS: - get_phone_number: inspect one returned phone_number_id in detail - list_agents: cross-check the agents assigned to a phone num

• make_outbound_call

  Place one outbound call through an ElevenLabs phone number. WHEN TO USE: - Call one recipient immediately after confirming the phone number assignment - Validate a telephony setup before creating a larger scheduled batch EXAMPLE: {"agent_id": "agent_123", "phone_number_id": "pn_123", "to_number": "+14155559876", "dynamic_variables": {"customer_name": "Jane"}} RELATED TOOLS: - get_phone_number: ins

• open_help_center

  Open the help center in the app.

• rebuild_knowledge_base_rag_index

  Trigger RAG indexing for one knowledge-base document (or read back its current index status when already indexed). WHEN TO USE: - Right after add_knowledge_base_document, so retrieval is ready before an agent is tested - When get_knowledge_base_rag_index_status shows a missing or failed index EXAMPLE: {"documentation_id": "doc_123"} RELATED TOOLS: - get_knowledge_base_rag_index_status: poll the st

• retry_batch_call

  Retry a previously submitted batch-call job. WHEN TO USE: - Re-run a batch after fixing quota, provider, or recipient issues - Create a fresh retry pass for failed or partial recipients EXAMPLE: {"batch_id": "batch_123"} RELATED TOOLS: - get_batch_call: inspect which recipients failed before retrying - list_batch_calls: find the correct batch_id - submit_batch_call: create a new batch instead when

• simulate_conversationwrite action

  Run a simulated conversation between one agent and a simulated user. This burns LLM credits but does not place a real call. WHEN TO USE: - Test prompt, voice, and tool-routing changes before telephony work - Validate a draft or duplicate agent with one short starter message EXAMPLE: {"agent_id": "agent_123", "user_message": "I need help rescheduling my appointment."} RELATED TOOLS: - update_agent:

• submit_batch_callwrite action
• submit_conversation_feedbackwrite action

  Submit like/dislike feedback for one conversation, closing the quality-review loop. WHEN TO USE: - Mark a conversation as good or bad after reviewing its transcript - Flag a failed call so it is visible in ElevenLabs analytics EXAMPLE: {"conversation_id": "conv_123", "feedback": "dislike"} RELATED TOOLS: - get_conversation: review the transcript and analysis before judging - list_conversations: di

• update_agentwrite action
• update_phone_numberwrite action

  Update the label and/or assigned agent on one ElevenLabs phone number. WHEN TO USE: - Assign a different agent before outbound or batch calling - Rename a phone number label so future telephony work is easier to identify EXAMPLE: {"phone_number_id": "pn_123", "label": "Sales line", "agent_id": "agent_123"} RELATED TOOLS: - get_phone_number: inspect the current label and assignment first - make_out

• Code scan51 source files scanned25/25
• –Live reliabilityno gateway calls yet and no remote to proben/a
• –Tool poisoningtools not inspected (local package is not executed); not countedn/a
• Auth qualitystatic API keys via environment variables6/15
• Maintenancelast push 0 days ago15/15
• Maintainer identityregistry namespace matches repository owner; GitHub account older than a year9/10

Safety score

85/100

higher than 72% of 48,114 servers

Tools

33

the typical server has 7

Tools that change things

33%

typical: 0%. Fewer write tools means less an agent can break

GitHub stars

14

more than 89% of 28,851 servers

Downloads a week

68

more than 42% of 15,093 servers with a package

ElevenLabs Agents: common questions

Is ElevenLabs Agents MCP server safe?

Yes, by our scan: it is graded A (85/100). Read the ElevenLabs Agents safety report

How do I install ElevenLabs Agents?

It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.

Does ElevenLabs Agents need an API key?

Yes. The registry entry asks for ELEVENLABS_API_KEY.

Is ElevenLabs Agents maintained?

The last commit was 11 days ago (2026-11-04). The latest release is v0.2.1.

How does ElevenLabs Agents compare with other MCP servers?

safety score 85/100 (higher than 72% of 48,114 servers); tools 33 (the typical server has 7); tools that change things 33% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.