Choose your preferred MCP client
Run this command in your terminal:
claude mcp add --transport http "agentphone" https://mcp.agentphone.ai/mcp
Click the button below to add this MCP server to Cursor:
Open in CursorOr add manually: Settings → MCP → Add server
Click the button below to add this MCP server to VS Code:
Open in VS CodeOr add manually: Settings → MCP → Add server
Click the button below to add this MCP server to VS Code Insiders:
Open in VS Code InsidersOr add manually: Settings → MCP → Add server
account_overviewGet a complete snapshot of your AgentPhone account: agents, phone numbers, webhook status, and usage limits.
list_numbersList all phone numbers in your account. Returns each number with its ID, status, country, and the agent it is assigned to (if any).
buy_numberPurchase a new phone number. This is a paid purchase that charges the account's balance, and requires confirm=true to actually go through. Returns the purchased number's details: its phone number, ID, country, and status.
send_messageSend an SMS, iMessage, or WhatsApp message from one of your agent's phone numbers, or react to an existing message. Each message is individually authorized by the user: this returns a preview and sends nothing unless confirm=true, so confirm the recipient(s) and content with the user before sending. The tool has two modes, so which inputs are required depends on the mode: - Send: give a destination and a sender (agent_id, from_number, or number_id), plus body and/or media. The destination is either to_number (a single recipient or existing group) or recipients (members of a NEW iMessage group) — exactly one. The agent must have a phone number attached; number_id or from_number picks which one when it has several. - React: set reaction and react_to_message_id (iMessage only) to tapback or emoji-react to a message. Pass no send fields; those are rejected. Channel: omit channel to auto-pick SMS/iMessage; set channel=whatsapp to force WhatsApp, which also unlocks the interactive fields (buttons, list, cta, template — one per message, and a template is needed to start a new WhatsApp conversation). iMessage extras (ignored on SMS): reply_to_message_id threads a reply, and send_style adds an expressive effect. Returns the sent message (id, status, channel, from/to numbers), or the applied reaction.
check_capabilitiesCheck which messaging channels a phone number can receive before you send: iMessage, SMS, and RCS. Use this to pick a channel or decide whether a number is reachable. Returns the per-channel flags and when they were last checked.
check_readinessCheck whether a phone number is ready to send a message or place a call BEFORE you try, so you can fix predictable failures first. Verifies the number is active, assigned to an agent, and capable of the channel you intend, plus the account's A2P/10DLC registration (for US SMS). Optionally checks a recipient's reachability. Returns a ready flag with specific blockers, warnings, and next steps. Note: credit balance and concurrency limits are enforced by the server at send time and cannot be pre-checked here, so a ready result does not guarantee against an insufficient-credit or busy error.
watch_messagesWatch one or more of your numbers for inbound messages (SMS, iMessage, or WhatsApp), identified by their number IDs. Use it to wait for a reply instead of calling get_messages repeatedly. To watch several numbers, pass them all in one call rather than calling once per number. A watch lasts until 24h pass with no new message on that number; calling again for the same number extends it. Returns once the watches are set up; when a message arrives, read it with get_messages.
get_messagesGet SMS messages for a specific phone number, identified by its number ID. Returns a list of individual messages — each with its timestamp, sender, recipient, and text — rather than grouped conversation threads. Results are newest first. To page through older history, pass the nextCursor from one response as cursor on the next; before/after are optional time filters only.
list_callsList recent calls. Choose ONE scope: the whole account (default), a single agent_id, or a single number_id. The status, direction, and search filters are only supported on the account-wide list. They cannot be combined with agent_id or number_id, and passing them together is rejected rather than silently ignored. The response reports the scope and the filters actually applied. Returns call summaries including each call's ID, direction, participants, and status.
get_callGet details and transcript for a specific call, identified by its call ID. Returns the call's direction, participants, status, timestamps, and transcript so far, as of now. A call that is still in progress returns its current state; call get_call again to see later updates.
end_callEnd an in-progress call immediately, identified by its call ID. Use this to hang up a call you started (or any active call in your account). Ending a call that has already finished is a no-op, so it is safe to retry. Returns the call's resulting status.
make_callInitiate an outbound, webhook-driven phone call where your backend handles the conversation logic. Each call is individually authorized by the user: this returns a preview and places no call unless confirm=true, so confirm the destination with the user before dialing. The agent must have a phone number attached and a webhook configured. This MCP connector prepends an automated-assistant disclosure to the greeting sent to the API. The one exception is wait_for_speaker: it connects silently and sends no greeting, so your call logic must have the agent identify as an AI when it first speaks. Calls made directly through the AgentPhone API or SDKs bypass this connector's disclosure logic. Returns the new call's ID, status, and the numbers it connects.
make_conversation_callPlace a phone call where the AI has an autonomous conversation about a given topic — scheduling, surveys, follow-ups, etc. No webhook setup needed. Each call is individually authorized by the user: this returns a preview and places no call unless confirm=true. Confirm the destination and topic with the user before dialing. The agent must have a phone number attached. This MCP connector prepends an automated-assistant disclosure to the greeting sent to the API and adds AI identity instructions around the topic. wait_for_speaker skips the spoken greeting prefix (the call waits for the callee), but the AI identity instructions still apply; no argument removes those. These guide the model but do not guarantee its behavior. Calls made directly through the AgentPhone API or SDKs bypass this connector's disclosure logic. This returns immediately with the call's ID and initial status. The call then runs on its own; use get_call with the call ID to fetch the transcript once it has progressed.
list_agentsList agents in your account, one page at a time. An agent is required before you can make calls — it owns phone numbers and handles voice/SMS. Returns each agent's ID, name, voice configuration, and attached phone numbers. Use offset to page through accounts with more agents than fit in one page; the response reports total and whether more remain.
create_agentCreate a new agent. An agent owns phone numbers and handles calls/SMS. A newly created agent has no phone number until one is attached. Set voice_mode to 'hosted' with a system_prompt for autonomous AI voice calls, or 'webhook' (default) to forward call transcripts to your webhook URL. Returns the created agent, including its new ID and full configuration.
update_agentUpdate an agent's configuration — name, description, voice settings, instructions, greeting, call transfer, or voicemail. Only provided fields are updated; omitted fields keep their current value. Switching voice_mode to 'hosted' requires a system_prompt. Returns the agent's full configuration after the update.
delete_agentDelete an agent permanently. Phone numbers attached to it are kept but unassigned. This cannot be undone. Returns the deleted agent's name and ID as confirmation.
get_agentGet one agent's full configuration. Returns its voice settings, system prompt, model tier, transfer and voicemail setup, and the phone numbers attached to it.
attach_numberAttach a phone number to an agent so the agent handles calls/SMS on that number. Takes an agent ID and a phone number ID. Returns the attached number and the agent it now belongs to.
detach_numberDetach a phone number from an agent. The number is kept in your account but becomes unassigned, and the agent stops handling calls and texts on it.
list_voicesList the voices available for agents. Returns each voice's ID (used in an agent's voice field), name, provider, gender, and accent.
list_conversationsList SMS conversations. Optionally filter by agent_id to see conversations for a specific agent. Each conversation is a thread between your number and an external contact, with its ID, participant, message count, and last-message preview.
get_conversationGet one SMS conversation with its message history (up to message_limit most recent messages), identified by its conversation ID. Returns the contact, your phone number, the message count, any stored metadata, and the messages.
get_conversation_messagesPage through a conversation's full message history, identified by its conversation ID. Unlike get_conversation (which returns the conversation detail with only a bounded tail of recent messages), this reaches arbitrarily far back: pass the nextCursor from one response as cursor on the next. Results are newest first; before/after are optional time filters only.
update_conversationSet metadata on a conversation. Use this to store custom state, tags, or context that persists between messages. Pass null to clear metadata.
list_contactsList saved contacts (your address book). Returns each contact's name, phone number, email, and ID. Optionally filter with a search term that matches name or phone number.
manage_contactCreate, update, or delete a saved contact. Set `action` to choose the operation. - create: requires phone_number and name - update: requires contact_id; only the fields you pass are changed - delete: requires contact_id, and confirm=true — permanent and cannot be undone. Without confirm the tool deletes nothing and returns a preview of what would be removed, so the user can approve first. create and update return the saved contact (id, name, phone number, email, notes); delete returns a success flag.
get_usageGet account usage statistics. By default returns a summary with number usage plus message, call, and webhook volume. Use breakdown='daily' or 'monthly' for time-series data.
get_webhookGet the webhook configuration — its URL and status. The signing secret is never returned (it would land in model context, transcripts and logs); retrieve it from the AgentPhone dashboard. Pass agent_id to get an agent-specific webhook, or omit for the project-level default.
set_webhookSet a webhook URL to receive inbound messages and call events. Pass agent_id to set a webhook for a specific agent (overrides project default). Omit agent_id to set the project-level webhook for all agents. The signing secret is never returned here (it would land in model context, transcripts and logs); retrieve it from the AgentPhone dashboard to verify webhook signatures.
delete_webhookRemove a webhook. Deleting an agent's webhook makes that agent fall back to the project-level webhook; deleting the project-level webhook stops events for agents using that default until you set one again (agents with their own webhook are unaffected). Pass agent_id to remove an agent's webhook, or omit it to remove the project-level webhook.
test_webhookSend a test event to a configured webhook and report whether your endpoint accepted it. Returns success, the HTTP status your endpoint returned, and any error message. Pass agent_id to test that agent's webhook. Omit to test the project-level webhook. Each call dispatches a real event to your endpoint, so it is not a no-op to repeat.
list_webhook_deliveriesView recent webhook delivery history. Shows which events were delivered, HTTP status codes, and timing. Pass agent_id to see deliveries for that agent's webhook. Omit for project-level. The hours time filter is supported only for project-level history.