← The evidence ledger
Evidence row 221994
Observed , by probe.mcp/1.
- Kind
- mcp.tools
- About
- endpoint b028a026-1a23-4a6c-99ca-321c7d49e6a7
- Recorded
- Previous row
- None: the first of its kind about this subject
- Body SHA-256
- aa604e5f8c71ee0ffff374cde1950b1c6811e3d3ff9323cd76f3d75ef87e1b6b
- Leaf
- 0x1dc76391c8c3aab108c26b2f979bb1a52e95bcc01742566bc0937aad756db537
What was observed
{
"count": 38,
"tools": [
{
"description": "Buys a real phone number. It charges the setup fee and the first\n month now and renews monthly. Call quote_numbers first. Pass the\n offer's quote_id, country_code, number_type, and confirm_total_cents =\n monthly_cents + setup_cents from that offer. If the quote expired or\n the price changed (409) you get an error: quote again. Only buy when\n your user asked for this number. After buying, call route_number to\n point it at an agent.\n\n A quote buys at most one number. After a timeout, retry with the same\n quote_id and idempotency_key, or check list_numbers. Never request a\n new quote after a timeout.\n\n ``idempotency_key`` is optional; to retry after a timeout or a\n 5xx, pass the ``idempotency_key`` that result returned, so the number\n is bought once. A 4xx error returns no key: the API remembers that\n failure under it, so drop it and quote again.\n\n ``verification_id`` picks which approved verification buys a number\n that needs one: the ``id`` of an approved entry in the offer's\n ``verifications``. Pass it when more than one is approved: show the\n user each ``subject_name`` (the person's or company's name) and ask\n which. With one approved, it is used. ``subject_type`` (``person`` or\n ``business``) is optional: it keeps only that kind. A 422 on\n ``verification_id`` means several are approved and none was picked,\n or the id is not an approved verification for this number.\n\n Returns the number (``id``, ``e164``, ``provisioning_state``,\n ``verification_id``, ``subject_name``, ...)\n plus ``idempotency_key``, or ``{\"error\": \"<message>\",\n \"idempotency_key\": ...}``.",
"inputSchema": {
"properties": {
"confirm_total_cents": {
"title": "Confirm Total Cents",
"type": "integer"
},
"country_code": {
"title": "Country Code",
"type": "string"
},
"idempotency_key": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Idempotency Key"
},
"number_type": {
"title": "Number Type",
"type": "string"
},
"quote_id": {
"title": "Quote Id",
"type": "string"
},
"subject_type": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Subject Type"
},
"verification_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Verification Id"
}
},
"required": [
"quote_id",
"country_code",
"number_type",
"confirm_total_cents"
],
"title": "acquire_number_toolArguments",
"type": "object"
},
"name": "acquire_number"
},
{
"description": "Cancel a verification that has not been sent for review, or dismiss\n a rejected one (needed before submitting a new one for the same\n country and number type). The draft is discarded.\n\n Works in state ``awaiting_review`` or ``rejected``, and for an\n ``approved`` US SMS registration (``purpose: \"sms\"``): its numbers\n leave the campaign, US-to-US texts from them fail again, and its\n monthly fee stops after the current month. Some registrations bill at\n least 3 monthly fees from approval, so cancelling sooner can still\n charge 3 months. Ask the user first. A\n ``submitting`` or ``submitted`` verification is under review and\n cannot be cancelled (409): wait for the result with get_verification.\n\n Returns the verification with ``state: \"cancelled\"``, or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"verification_id": {
"title": "Verification Id",
"type": "string"
}
},
"required": [
"verification_id"
],
"title": "cancel_verification_toolArguments",
"type": "object"
},
"name": "cancel_verification"
},
{
"description": "When a registration shows confirmation 'code', ask the person for\n the code the carrier texted and submit it here.\n\n ``verification_id`` comes from submit_verification or\n list_verifications. ``code`` is 4 to 10 characters. A wrong or expired\n code is an error naming ``code``; the carrier sends a new one after a\n day. On success the registration moves on; poll get_verification\n until ``approved``. A ``confirmation`` of ``reply`` means the person\n answers the carrier's text instead; nothing is entered here.\n\n Returns the verification or ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"code": {
"title": "Code",
"type": "string"
},
"verification_id": {
"title": "Verification Id",
"type": "string"
}
},
"required": [
"verification_id",
"code"
],
"title": "confirm_verification_toolArguments",
"type": "object"
},
"name": "confirm_verification"
},
{
"description": "Save an agent. Then call ``route_number`` so a number answers\n with it, or pass its id as ``agent_id`` to ``place_call``.\n\n ``system_prompt`` is what the agent follows. ``first_message`` is\n spoken after the AI line; omit it to let the caller speak first.\n ``ai_disclosure_line`` replaces the workspace AI line for this\n agent; ``{org}`` becomes the organization name. ``language`` is a\n lowercase ISO 639-1 code. ``max_duration_seconds`` is 60..3600.\n ``handover_max_duration_seconds`` (60..3600, default 1800) is the\n call limit once a handover contact answers.\n ``handover_contacts``: People the agent may hand a live call to:\n ``[{contact_id, note}]`` (max 10; ``note`` 1-200 chars says when to\n hand over). ``contact_id`` is an id from ``list_contacts``: a\n contact's uuid, or ``member:<user id>`` for a team member. The\n person needs a phone number. The agent never sees it.\n\n Returns the API's ``AgentResponse`` as a dict, or\n ``{\"error\": \"<message>\"}``.\n ",
"inputSchema": {
"properties": {
"ai_disclosure": {
"default": true,
"title": "Ai Disclosure",
"type": "boolean"
},
"ai_disclosure_line": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Ai Disclosure Line"
},
"first_message": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "First Message"
},
"handover_contacts": {
"anyOf": [
{
"items": {
"additionalProperties": true,
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Handover Contacts"
},
"handover_max_duration_seconds": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"title": "Handover Max Duration Seconds"
},
"language": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Language"
},
"max_duration_seconds": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"title": "Max Duration Seconds"
},
"name": {
"title": "Name",
"type": "string"
},
"recording_enabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"title": "Recording Enabled"
},
"sms_enabled": {
"default": true,
"title": "Sms Enabled",
"type": "boolean"
},
"system_prompt": {
"title": "System Prompt",
"type": "string"
},
"tools": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Tools"
},
"voice_enabled": {
"default": true,
"title": "Voice Enabled",
"type": "boolean"
},
"voice_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Voice Id"
}
},
"required": [
"name",
"system_prompt"
],
"title": "create_agent_toolArguments",
"type": "object"
},
"name": "create_agent"
},
{
"description": "Save a new contact for the workspace. Provide at least one of\n ``phone_e164`` (E.164, e.g. ``+14155551234``) or ``email`` — the API\n rejects (422) a contact with neither. A duplicate phone or email on\n an existing contact returns 409.\n\n Example:\n create_contact(name=\"Maya Chen\", phone_e164=\"+14155551234\")\n\n Returns the created contact entry as a dict. On failure returns\n ``{\"error\": \"<message>\"}`` instead.\n ",
"inputSchema": {
"properties": {
"email": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Email"
},
"name": {
"title": "Name",
"type": "string"
},
"phone_e164": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Phone E164"
}
},
"required": [
"name"
],
"title": "create_contact_toolArguments",
"type": "object"
},
"name": "create_contact"
},
{
"description": "Delete a saved agent. DESTRUCTIVE. Ask the user before calling this.\n\n Numbers that used the agent stop answering with it: calls ring out\n and texts reach webhooks only. Reroute them first with\n ``route_number`` if that matters.\n\n ``confirm_name`` must equal the agent's ``name``. The tool reads the\n agent first and refuses with ``{\"error\": ...}`` on a mismatch,\n without deleting anything. To stop an agent without deleting it, use\n ``update_agent`` with ``status=\"paused\"``.\n\n Returns ``{\"deleted\": true, \"agent_id\", \"name\"}``, or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"agent_id": {
"title": "Agent Id",
"type": "string"
},
"confirm_name": {
"title": "Confirm Name",
"type": "string"
}
},
"required": [
"agent_id",
"confirm_name"
],
"title": "delete_agent_toolArguments",
"type": "object"
},
"name": "delete_agent"
},
{
"description": "Fetch one saved agent by id. ``agent_id`` comes from ``list_agents``\n or ``create_agent``.\n\n Returns the ``AgentResponse`` as a dict (``name``, ``system_prompt``,\n ``first_message``, ``ai_disclosure``, ``voice_config``, ``tools``,\n ``voice_enabled``, ``sms_enabled``, ``status``, ...), or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"agent_id": {
"title": "Agent Id",
"type": "string"
}
},
"required": [
"agent_id"
],
"title": "get_agent_toolArguments",
"type": "object"
},
"name": "get_agent"
},
{
"description": "Fetch the current state of one call by id.\n\n Use this after ``place_call`` (or to check on any prior call)\n to read the call's latest ``status`` and timing fields.\n\n Returns the API's ``CallResponse`` as a dict, or\n ``{\"error\": \"call not found\"}`` for an unknown id.\n ",
"inputSchema": {
"properties": {
"call_id": {
"title": "Call Id",
"type": "string"
}
},
"required": [
"call_id"
],
"title": "get_call_toolArguments",
"type": "object"
},
"name": "get_call"
},
{
"description": "Fetch the full record of one email by id.\n\n Returns the complete row — including ``body_text`` / ``body_html``\n and (for inbound mail) the ``in_reply_to`` / ``message_id`` headers\n and ``spam``/``virus``/``spf``/``dkim``/``dmarc`` verdicts. Use this\n after ``list_emails`` to read a received reply's body.\n\n Returns the API's ``EmailResponse`` as a dict, or\n ``{\"error\": \"resource not found\"}`` for an unknown id.\n ",
"inputSchema": {
"properties": {
"email_id": {
"title": "Email Id",
"type": "string"
}
},
"required": [
"email_id"
],
"title": "get_email_toolArguments",
"type": "object"
},
"name": "get_email"
},
{
"description": "Get a fetchable URL for one inbound email attachment.\n\n ``attachment_id`` comes from an item in ``get_email``'s\n ``attachments`` list. Returns ``{\"url\": \"<presigned-s3-url>\"}`` —\n a short-lived (~5 minute) link to the attachment bytes, fetchable\n directly with no auth header. Unknown ids return\n ``{\"error\": \"resource not found\"}``.\n ",
"inputSchema": {
"properties": {
"attachment_id": {
"title": "Attachment Id",
"type": "string"
},
"email_id": {
"title": "Email Id",
"type": "string"
}
},
"required": [
"email_id",
"attachment_id"
],
"title": "get_email_attachment_toolArguments",
"type": "object"
},
"name": "get_email_attachment"
},
{
"description": "Delivery/engagement timeline (sent→delivered→opened…) for one email.\n\n Chronological lifecycle events for a single email — use this to see\n exactly what happened to one message (bounced? opened? clicked?)\n rather than the account-wide aggregates ``get_email_stats`` returns.\n\n Returns ``{\"items\": [...], \"next_cursor\": ...}`` where each item has\n ``kind`` (one of sent, delivered, delivery_delayed, bounced,\n complained, rejected, opened, clicked), ``payload``, and\n ``occurred_at``. ``opened`` and ``clicked`` only occur for emails\n sent with an HTML body. A plain-text-only email still gets sent,\n delivered, and bounce events, but opens and clicks are never\n tracked. Pass ``cursor`` from a previous ``next_cursor`` to\n page (``limit`` 1..1000, default 100). On failure returns\n ``{\"error\": \"resource not found\"}`` for an unknown id.\n ",
"inputSchema": {
"properties": {
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Cursor"
},
"email_id": {
"title": "Email Id",
"type": "string"
},
"limit": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"title": "Limit"
}
},
"required": [
"email_id"
],
"title": "get_email_events_toolArguments",
"type": "object"
},
"name": "get_email_events"
},
{
"description": "Get a fetchable URL for an email's original MIME source.\n\n Returns ``{\"url\": \"<presigned-s3-url>\"}`` — a short-lived\n (~5 minute) link to the full raw RFC822 message. Fetch the URL\n directly to read the complete original; it needs no auth header.\n Raw source exists for **inbound** mail only — outbound ids return\n ``{\"error\": \"resource not found\"}``.\n ",
"inputSchema": {
"properties": {
"email_id": {
"title": "Email Id",
"type": "string"
}
},
"required": [
"email_id"
],
"title": "get_email_raw_toolArguments",
"type": "object"
},
"name": "get_email_raw"
},
{
"description": "Account-level email deliverability stats (counts, rates, time series).\n\n Aggregates across your whole organization's outbound mail over a\n window — use this for \"how's deliverability doing\" rather than\n one message's history (``get_email_events`` covers that).\n\n ``from_`` / ``to`` are ISO 8601 timestamps (defaults: last 7 days\n ending now). ``bucket`` is ``\"day\"`` (default) or ``\"hour\"`` —\n ``\"hour\"`` is limited to an 8-day span; any range is capped at 92\n days.\n\n Returns ``{\"from\": ..., \"to\": ..., \"bucket\": ..., \"totals\": {...},\n \"rates\": {...}, \"series\": [...]}`` — ``totals``/each ``series``\n bucket carry counts (sent, delivered, bounced, opened, ...) and\n ``rates`` carries derived ratios (delivery, bounce, open, click),\n each ``None`` when ``totals.sent`` is 0. Open and click numbers\n only count HTML emails. Plain-text-only emails are never tracked\n for opens or clicks, so they lower those rates. On failure returns\n ``{\"error\": \"<message>\"}`` instead.\n ",
"inputSchema": {
"properties": {
"bucket": {
"default": "day",
"title": "Bucket",
"type": "string"
},
"from_": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "From"
},
"to": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "To"
}
},
"title": "get_email_stats_toolArguments",
"type": "object"
},
"name": "get_email_stats"
},
{
"description": "Page through events from across the org or one resource.\n\n Pass ``id=\"<type>:<uuid>\"`` to narrow to a single resource —\n supported types are ``call``, ``email``, and ``sms`` (e.g.\n ``id=\"sms:<uuid>\"`` after ``send_sms``). When narrowed to a\n call, the response includes a ``call_status`` field reflecting\n the call's current state. Without ``id``, returns events from\n across the whole org. ``kind`` filters server-side by event kind\n (``state_change``, ``agent_turn``, ``user_turn``, ``tool_call``,\n ``error``, ...).\n\n This is **not** a streaming subscription — the call returns\n whatever events exist now plus a ``next_cursor`` if more pages\n remain. To follow a call to completion, loop: pass the previous\n response's ``next_cursor`` until ``next_cursor`` is null and,\n when narrowed to a call, ``call_status`` is one of\n ``completed``, ``failed``, ``busy``, ``no_answer``, ``canceled``\n (the terminal set).\n\n Example:\n get_events(id=\"call:0c2f...-...\", limit=200)\n\n Returns ``{\"items\": [...], \"next_cursor\": <str|None>,\n \"call_status\": <str|None>}`` on success, or\n ``{\"error\": ...}`` on a malformed ``id`` or upstream failure.\n ",
"inputSchema": {
"properties": {
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Cursor"
},
"id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Id"
},
"kind": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Kind"
},
"limit": {
"default": 100,
"title": "Limit",
"type": "integer"
}
},
"title": "get_events_toolArguments",
"type": "object"
},
"name": "get_events"
},
{
"description": "Fetch one dedicated number by id: its state, capabilities and which\n agents answer its calls and texts. ``number_id`` comes from\n ``list_numbers``.\n\n Returns the ``PhoneNumberResponse`` as a dict, or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"number_id": {
"title": "Number Id",
"type": "string"
}
},
"required": [
"number_id"
],
"title": "get_number_toolArguments",
"type": "object"
},
"name": "get_number"
},
{
"description": "Fetch the current state of one SMS by id.\n\n Use this after ``send_sms`` to check delivery status.\n\n Example:\n get_sms(sms_id=\"...\")\n ",
"inputSchema": {
"properties": {
"sms_id": {
"title": "Sms Id",
"type": "string"
}
},
"required": [
"sms_id"
],
"title": "get_sms_toolArguments",
"type": "object"
},
"name": "get_sms"
},
{
"description": "Fetch one verification and its state. Use it to poll.\n\n There is no notification, and review takes business days. Poll every\n few minutes to hours, not every few seconds; the state refreshes from\n the carrier at most once a minute.\n\n On ``approved``: call quote_numbers again. The offer is now\n ``ready``; then acquire_number. On ``rejected``: read\n ``rejection_reason``, ask the user, dismiss it with\n cancel_verification, then submit a new one. ``verification_id`` comes\n from submit_verification or list_verifications.\n\n Returns ``{\"id\", \"state\", \"rejection_reason\", ...}`` or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"verification_id": {
"title": "Verification Id",
"type": "string"
}
},
"required": [
"verification_id"
],
"title": "get_verification_toolArguments",
"type": "object"
},
"name": "get_verification"
},
{
"description": "Read what the carrier needs to verify the organization before it\n can buy a number. Call this when quote_numbers shows an offer with\n ``readiness == \"verification_required\"``. Read-only.\n\n ``country_code`` is ISO alpha-2 (for example ``\"GB\"``). ``number_type``\n is ``local``, ``mobile``, ``national`` or ``toll_free``.\n ``subject_type`` is required: ``person`` (the user) or ``business``\n (the user's company). Never guess it: ask the user \"you or your\n company?\". The result's ``subject_types`` lists what the carrier\n accepts. The organization may hold one verification of each. ``quote_id`` is the\n offer's ``quote_id`` from quote_numbers. It is required for\n ``purpose='number'``: the form is the one for that offer.\n\n ``purpose`` is ``number`` (default, to buy a number) or ``sms``.\n ``purpose='sms'`` registers the organization to text US numbers from a\n US number (country_code US, number_type local). The carrier fee is\n debited from credit at filing. If the error says to pass\n ``number_id``, pass the ``id`` of one of the organization's US numbers\n from list_numbers.\n\n Returns ``required`` (false means no verification is needed and\n submit_verification would fail), ``fields`` (values to collect),\n ``documents`` (slots, each with ``options``), ``address_required``.\n Skip any document slot whose ``needs_input`` is false: do not send it.\n\n Ask the USER for every value and every file. Never invent, guess or\n infer personal or company data. Never reuse data from another task.\n\n Errors come back as ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"country_code": {
"title": "Country Code",
"type": "string"
},
"number_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Number Id"
},
"number_type": {
"title": "Number Type",
"type": "string"
},
"purpose": {
"default": "number",
"title": "Purpose",
"type": "string"
},
"quote_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Quote Id"
},
"subject_type": {
"title": "Subject Type",
"type": "string"
}
},
"required": [
"country_code",
"number_type",
"subject_type"
],
"title": "get_verification_requirements_toolArguments",
"type": "object"
},
"name": "get_verification_requirements"
},
{
"description": "List the workspace's saved agents: the brains that answer calls\n and texts on its numbers, and that ``place_call`` can use via\n ``agent_id``.\n\n Returns ``{\"items\": [{\"id\", \"name\", \"system_prompt\", \"first_message\",\n \"ai_disclosure\", \"ai_disclosure_line\", \"voice_config\", \"tools\",\n \"max_duration_seconds\", \"handover_max_duration_seconds\", \"voice_enabled\",\n \"sms_enabled\", \"status\", \"handover_contacts\"}, ...]}``.\n ",
"inputSchema": {
"properties": {},
"title": "list_agents_toolArguments",
"type": "object"
},
"name": "list_agents"
},
{
"description": "List recent calls in your organization, newest first.\n\n Cursor-paginated: pass the previous response's ``next_cursor``\n to fetch the next page. ``status`` (one of queued, dialing,\n ringing, in_progress, completed, failed, busy, no_answer,\n canceled) and ``to`` (E.164) are optional server-side filters.\n\n Returns a dict ``{\"items\": [...], \"next_cursor\": <str|None>}``.\n ",
"inputSchema": {
"properties": {
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Cursor"
},
"limit": {
"default": 50,
"title": "Limit",
"type": "integer"
},
"status": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Status"
},
"to": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "To"
}
},
"title": "list_calls_toolArguments",
"type": "object"
},
"name": "list_calls"
},
{
"description": "List the workspace's contacts: org members (with their phone/email)\n plus manually saved contacts. Use ``lookup_contact`` for name searches.\n\n ``q`` optionally filters server-side (matches name/email/phone).\n ``limit`` caps the page (server default 100, max 500).\n\n Returns ``{\"items\": [{\"id\", \"kind\", \"name\", \"phone_e164\", \"email\",\n \"role\"}, ...]}`` — ``kind`` is ``\"member\"`` (id ``member:<user_id>``,\n ``role`` set) or ``\"manual\"`` (id is the contact's UUID, ``role``\n null).\n ",
"inputSchema": {
"properties": {
"limit": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"title": "Limit"
},
"q": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Q"
}
},
"title": "list_contacts_toolArguments",
"type": "object"
},
"name": "list_contacts"
},
{
"description": "List the addresses this workspace can send email from. Call this\n BEFORE ``send_email`` when you do not already know the ``from_``\n address — do not guess a domain.\n\n Returns ``{\"items\": [...], \"next_cursor\": ..., \"default_from\": ...}``.\n Each item carries ``domain``, ``kind`` (``\"custom\"`` or\n ``\"hail_mail\"``) and ``verification_status``; only ``\"verified\"``\n rows can send. A ``custom`` item's ``domain`` is a bare DNS name\n that accepts any local-part (``sales@acme.com``); a ``hail_mail``\n item's ``domain`` is already a full address.\n\n ``default_from`` is the address a ``send_email`` with no ``from_``\n goes out as. It is null when the workspace has several verified\n identities — then ``send_email`` rejects a missing ``from_`` and you\n must pass one from ``items``.\n\n Example:\n list_email_domains()\n ",
"inputSchema": {
"properties": {
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Cursor"
},
"limit": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"title": "Limit"
}
},
"title": "list_email_domains_toolArguments",
"type": "object"
},
"name": "list_email_domains"
},
{
"description": "List emails in your organization, newest first.\n\n Cursor-paginated: pass the previous response's ``next_cursor`` to\n fetch the next page. Two optional server-side filters:\n\n * ``direction`` — ``outbound`` or ``inbound``. Pass\n ``direction=\"inbound\"`` to read replies and other received mail.\n * ``status`` — one of ``queued``, ``sent``, ``failed``,\n ``bounced``, ``complained``, ``received``.\n\n Items are trimmed summaries (no message body). Call ``get_email``\n with an item's ``id`` to read the full body.\n\n Example:\n list_emails(direction=\"inbound\")\n\n Returns ``{\"items\": [...], \"next_cursor\": <str|None>}`` on success,\n or ``{\"error\": \"<message>\"}`` instead.\n ",
"inputSchema": {
"properties": {
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Cursor"
},
"direction": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Direction"
},
"limit": {
"default": 50,
"title": "Limit",
"type": "integer"
},
"status": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Status"
}
},
"title": "list_emails_toolArguments",
"type": "object"
},
"name": "list_emails"
},
{
"description": "List the US state and Canadian province codes that quote_numbers\n accepts as ``region`` (optional, with ``city``).\n\n Returns ``{\"regions\": [{\"country_code\", \"code\", \"name\"}]}``, such as\n ``{\"country_code\": \"US\", \"code\": \"AZ\", \"name\": \"Arizona\"}``, or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {},
"title": "list_number_regions_toolArguments",
"type": "object"
},
"name": "list_number_regions"
},
{
"description": "List the workspace's dedicated phone numbers, newest first.\n\n ``limit`` is 1..200 (default 50). Pass the returned ``next_cursor``\n as ``cursor`` for the next page.\n\n Returns ``{\"items\": [{\"id\", \"e164\", \"country_code\",\n \"number_type\", \"capabilities\", \"provisioning_state\", \"voice_agent_id\",\n \"sms_agent_id\", ...}], \"next_cursor\"}``, or ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Cursor"
},
"limit": {
"default": 50,
"title": "Limit",
"type": "integer"
}
},
"title": "list_numbers_toolArguments",
"type": "object"
},
"name": "list_numbers"
},
{
"description": "Page through recent SMS messages for your organization.\n\n ``status`` filters to one of: queued, sent, delivered, failed,\n undelivered, received. ``to`` filters to messages sent to a\n specific E.164 number. Paginate with the returned\n ``next_cursor``.\n ",
"inputSchema": {
"properties": {
"cursor": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Cursor"
},
"limit": {
"default": 50,
"title": "Limit",
"type": "integer"
},
"status": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Status"
},
"to": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "To"
}
},
"title": "list_sms_toolArguments",
"type": "object"
},
"name": "list_sms"
},
{
"description": "List the organization's carrier verifications, newest first.\n\n ``purpose`` is ``number`` or ``sms``; only that kind is listed.\n Omit to see every registration. ``purpose='sms'`` lists the\n registrations to text US numbers from a US number.\n\n Returns ``{\"items\": [{\"id\", \"country_code\", \"number_type\",\n \"subject_type\", \"subject_name\", \"purpose\", \"state\",\n \"rejection_reason\", \"created_at\",\n \"updated_at\", \"submitted_at\", \"approved_at\", \"confirmation\"}]}``.\n ``state`` is ``awaiting_review``, ``submitting``, ``submitted``,\n ``approved``, ``rejected`` or ``cancelled``. Use it to check whether a\n submission already exists before submitting again. ``confirmation``\n ``code``: call confirm_verification with the code the carrier texted\n the person.\n\n Errors come back as ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"purpose": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Purpose"
}
},
"title": "list_verifications_toolArguments",
"type": "object"
},
"name": "list_verifications"
},
{
"description": "Find a contact by name, email, or phone fragment. Resolve a person\n to their ``phone_e164``/``email`` BEFORE calling ``place_call``,\n ``send_sms``, or ``send_email`` — do not guess a contact's number.\n\n Returns up to 10 matches, same item shape as ``list_contacts``.\n\n Example:\n lookup_contact(query=\"maya\")\n ",
"inputSchema": {
"properties": {
"query": {
"title": "Query",
"type": "string"
}
},
"required": [
"query"
],
"title": "lookup_contact_toolArguments",
"type": "object"
},
"name": "lookup_contact"
},
{
"description": "Originate an outbound phone call.\n\n Provide ``system_prompt`` (mode A — Hail's bundled fallback LLM\n uses this prompt) or ``llm`` (mode B — bring your own\n OpenAI-compatible endpoint as\n ``{\"base_url\": ..., \"api_key\": ..., \"model\": ...}``), or both, or\n ``agent_id`` (a saved agent from ``list_agents`` supplies the\n prompt, first message, AI line, voice and tools; fields you pass\n here win). At least one is required. Passing both runs your prompt on your\n own endpoint: it receives Hail's voice preamble plus your prompt\n as the leading system message.\n\n ``to`` must be E.164 (e.g. ``+14155551234``). ``from_`` is\n optional and defaults to the first active number on your org.\n ``first_message`` is spoken verbatim on pickup; omit it to let\n the agent open the conversation itself — it reacts to how the\n call was answered, or introduces itself after silence.\n ``language`` sets the call's spoken language for speech-to-text,\n text-to-speech, and turn detection, as a lowercase ISO 639-1 code\n (e.g. ``\"da\"``); 39 languages are supported (server rejects others\n with 422); omit for English. STT provider selection is\n console-BYO-only (no per-call override); configure it on the\n organization to pin a provider.\n ``ai_disclosure=False`` skips the spoken \"this is an AI\n assistant\" line at the start of the call. Leave enabled unless\n the user has verified it is not required for this call — US\n artificial-voice calls (47 CFR 64.1200(b)(1)) and several AI\n bot-disclosure laws require it, and Hail does not verify this.\n The agent still identifies itself as an AI if asked.\n ``recording_enabled``: omit it to follow the agent, then the\n workspace recording setting. ``true`` records this call and\n ``false`` does not, whatever the workspace setting. Pass it only\n when the user asked for or against a recording.\n ``metadata`` is free-form JSON attached to the call record.\n ``tools`` are the agent tools to allow on this call. Omit for all\n available; pass ``[]`` to disable.\n\n ``recipient_consent`` is required and must be the JSON boolean\n ``true`` (strings and numbers are refused): attest that you (the caller\n triggering this request) have obtained the lawful consent needed\n to contact this recipient. The API rejects the request (422) if\n this is not ``true`` — Hail does not verify consent for you, you\n are responsible for having a lawful basis (TCPA / ePrivacy / GDPR\n as applicable). Set ``message_type=\"marketing\"`` for promotional\n calls (this additionally requires a non-empty ``consent_source``\n describing how/where consent was obtained) — leave as the default\n ``\"informational\"`` for transactional/service calls.\n\n ``idempotency_key`` defaults to a fresh UUID per invocation\n and is returned in the response under ``idempotency_key`` — to\n retry *this* exact request (rather than dispatch a second call),\n pass the value back on the retry. A new key is a new call.\n\n Example:\n place_call(to=\"+14155551234\",\n recipient_consent=True,\n system_prompt=\"You are scheduling a haircut.\",\n first_message=\"Hi, I'm calling on behalf of Alex.\")\n\n Returns the API's ``CallResponse`` as a dict (id, status,\n from_e164, to_e164, ...). On failure returns\n ``{\"error\": \"<message>\"}`` instead.\n ",
"inputSchema": {
"properties": {
"agent_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Agent Id"
},
"ai_disclosure": {
"default": true,
"title": "Ai Disclosure",
"type": "boolean"
},
"consent_obtained_at": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Consent Obtained At"
},
"consent_source": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Consent Source"
},
"first_message": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "First Message"
},
"from_": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "From"
},
"idempotency_key": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Idempotency Key"
},
"language": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Language"
},
"llm": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"title": "Llm"
},
"message_type": {
"default": "informational",
"title": "Message Type",
"type": "string"
},
"metadata": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"title": "Metadata"
},
"recipient_consent": {
"title": "Recipient Consent",
"type": "boolean"
},
"recording_enabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"title": "Recording Enabled"
},
"system_prompt": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "System Prompt"
},
"to": {
"title": "To",
"type": "string"
},
"tools": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Tools"
}
},
"required": [
"to",
"recipient_consent"
],
"title": "place_call_toolArguments",
"type": "object"
},
"name": "place_call"
},
{
"description": "Compare live phone-number offers from the carriers. Read-only: it\n buys nothing and charges nothing.\n\n ``country_code`` is an ISO alpha-2 code such as ``\"US\"``.\n ``capabilities`` is ``[\"voice\"]``, ``[\"sms\"]`` or both (default).\n ``number_type`` is ``local``, ``mobile``, ``national`` or\n ``toll_free``; omit it to compare all types.\n ``area_code`` is an area code or the first digits of the number, as\n dialed in the country: ``\"020\"`` (London), ``\"480\"`` (Arizona).\n ``city`` is a city name such as ``\"London\"``. Each offer has\n ``locality`` and ``region`` (such as ``\"Phoenix\"``, ``\"AZ\"``): check\n them, since city names repeat across states. ``region`` (optional,\n with ``city``) is a US/CA state code; list_number_regions lists them.\n Omit all three to search the whole country.\n\n Returns ``{\"offers\": [...], \"recommended_quote_id\", \"expires_at\"}``.\n Each offer has ``e164``, ``number_type``,\n ``capabilities``, ``monthly_cents``, ``setup_cents``, ``readiness``,\n ``requirements``, ``subject_types``, ``verifications`` and\n ``quote_id``. Money fields are cents (USD).\n Quotes expire in 10 minutes (``expires_at``); quote again after that.\n\n ``subject_types`` says who may verify for the number: ``person``\n (the user), ``business`` (the user's company), or both; empty when no\n verification is needed. ``verifications`` lists the verifications\n the offer can be bought with (``id``, ``state``, ``subject_type``,\n ``subject_name``), approved first: those of an accepted subject type\n still under review, and the approved ones the carrier still accepts\n for the offer. ``subject_name`` is the person's or company's name from the\n verification form. When several are approved, show the user the\n names and ask which one buys the number; pass its ``id`` as\n ``verification_id`` to acquire_number.\n\n An offer with ``readiness == \"verification_required\"`` cannot be\n bought until the organization is verified. Ask the user \"you or your\n company?\" to pick the ``subject_type`` (one of ``subject_types``).\n Call get_verification_requirements, ask the user for the details,\n then submit_verification, both with the offer's ``quote_id`` and that\n ``subject_type``, and poll get_verification (review takes business\n days). ``requirements``\n lists what the carrier needs. Prefer an offer whose\n ``readiness`` is ``\"ready\"``; ``recommended_quote_id`` points at the\n best one, or is null if none is ready.\n\n An offer with ``by_request: true`` has no ``e164``: support orders\n that number by hand. acquire_number refuses it; tell the user to ask\n support. Its ``quote_id`` still works for verification.\n\n Errors come back as ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"area_code": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Area Code"
},
"capabilities": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Capabilities"
},
"city": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "City"
},
"country_code": {
"title": "Country Code",
"type": "string"
},
"number_type": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Number Type"
},
"region": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Region"
}
},
"required": [
"country_code"
],
"title": "quote_numbers_toolArguments",
"type": "object"
},
"name": "quote_numbers"
},
{
"description": "Release a phone number. DESTRUCTIVE and permanent.\n\n Calls and texts to the number stop. This month's fee and any earlier\n months are still billed. There is no refund. The number may not be\n recoverable.\n Ask the user before calling this.\n\n ``confirm_e164`` must equal the number's ``e164`` (for example\n ``\"+14155550100\"``). The tool reads the number first and refuses with\n ``{\"error\": ...}`` on a mismatch, without releasing anything. A number\n whose order is still pending cannot be released (409). A number whose\n order failed was never bought: the call only hides it from the list\n and returns ``{\"dismissed\": true, \"number_id\", \"e164\"}``.\n\n Returns ``{\"released\": true, \"number_id\", \"e164\"}``, or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"confirm_e164": {
"title": "Confirm E164",
"type": "string"
},
"number_id": {
"title": "Number Id",
"type": "string"
}
},
"required": [
"number_id",
"confirm_e164"
],
"title": "release_number_toolArguments",
"type": "object"
},
"name": "release_number"
},
{
"description": "Choose which agent answers a number's calls and texts.\n\n ``voice_agent_id`` makes the number answer calls with that agent\n (Hail registers it for inbound at the carrier and LiveKit);\n ``clear_voice=True`` stops answering calls. ``sms_agent_id`` makes\n the agent reply to texts; ``clear_sms=True`` stops that. A field\n left out keeps its value. ``number_id`` comes from GET /numbers.\n\n Returns the updated ``PhoneNumberResponse`` as a dict, or\n ``{\"error\": \"<message>\"}``.\n ",
"inputSchema": {
"properties": {
"clear_sms": {
"default": false,
"title": "Clear Sms",
"type": "boolean"
},
"clear_voice": {
"default": false,
"title": "Clear Voice",
"type": "boolean"
},
"number_id": {
"title": "Number Id",
"type": "string"
},
"sms_agent_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Sms Agent Id"
},
"voice_agent_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Voice Agent Id"
}
},
"required": [
"number_id"
],
"title": "route_number_toolArguments",
"type": "object"
},
"name": "route_number"
},
{
"description": "Send an outbound email through your configured SES sender.\n\n ``to`` is a non-empty list of RFC-style email addresses. At\n least one of ``body_text`` / ``body_html`` is required (both\n is fine — multipart-alternative). Prefer including\n ``body_html``: open and click tracking only works for emails\n with an HTML body. A plain-text-only email still gets sent,\n delivered, and bounce events, but opens and clicks are never\n tracked. ``cc``, ``bcc``, and\n ``reply_to`` are optional and follow the usual mail\n conventions. Hail sets no ``reply_to`` of its own — to have\n replies reach the human running this session, call ``whoami``\n and pass their ``email``.\n\n ``recipient_consent`` is required and must be the JSON boolean\n ``true`` (strings and numbers are refused): attest that you (the caller\n triggering this request) have obtained the lawful consent needed\n to email this recipient. The API rejects the request (422) if\n this is not ``true`` — Hail does not verify consent for you. Set\n ``message_type=\"marketing\"`` for promotional email (this\n additionally requires a non-empty ``consent_source`` describing\n how/where consent was obtained) — leave as the default\n ``\"informational\"`` for transactional/service email.\n\n ``from_`` is optional only while the workspace has one verified\n sender identity. With several, omitting it returns 422 listing\n them — call ``list_email_domains`` and pass one. With none, the\n server auto-mints a per-org hail-mail address of the form\n ``<user>+<org>@<base>`` if the operator configured\n ``HAIL_MAIL_BASE_DOMAIN`` (the ``<org>`` part is derived from your\n organization id; the ``<user>`` part comes from ``HAIL_MAIL_FROM``\n / ``HAIL_MAIL_DEFAULT_USER_PREFIX``, or an explicit row created\n via ``POST /email-domains``). When supplied, ``from_`` must match\n a verified row already in ``email_domains`` (register one with\n the website console or ``POST /email-domains``).\n\n ``from_name`` is an optional display name rendered on the\n From: header (\"Acme Billing <billing@acme.com>\"). Control\n characters are rejected.\n\n ``metadata`` is free-form JSON attached to the email record.\n\n ``idempotency_key`` defaults to a fresh UUID and is returned\n in the response under ``idempotency_key`` — pass the same\n value on a retry to replay rather than re-send. A new key\n is a new message.\n\n ``attachment_ids`` are ids returned by ``upload_email_attachment``\n — upload a file first, then pass its id(s) here to attach it.\n\n Example:\n send_email(to=[\"alice@example.com\"],\n subject=\"Welcome\",\n recipient_consent=True,\n body_text=\"Thanks for signing up.\")\n\n Returns the ``EmailResponse`` dict (id, status, from_address,\n to_addresses, sent_at, provider_message_id, ...). On failure\n returns ``{\"error\": \"<message>\"}`` instead.\n ",
"inputSchema": {
"properties": {
"attachment_ids": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Attachment Ids"
},
"bcc": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Bcc"
},
"body_html": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Body Html"
},
"body_text": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Body Text"
},
"cc": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Cc"
},
"consent_obtained_at": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Consent Obtained At"
},
"consent_source": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Consent Source"
},
"from_": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "From"
},
"from_name": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "From Name"
},
"idempotency_key": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Idempotency Key"
},
"message_type": {
"default": "informational",
"title": "Message Type",
"type": "string"
},
"metadata": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"title": "Metadata"
},
"recipient_consent": {
"title": "Recipient Consent",
"type": "boolean"
},
"reply_to": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Reply To"
},
"subject": {
"title": "Subject",
"type": "string"
},
"to": {
"items": {
"type": "string"
},
"title": "To",
"type": "array"
}
},
"required": [
"to",
"subject",
"recipient_consent"
],
"title": "send_email_toolArguments",
"type": "object"
},
"name": "send_email"
},
{
"description": "Send an outbound SMS.\n\n ``to`` must be E.164 (e.g. ``+14155551234``). ``body`` is the\n message text. With no ``from_``: destinations where the sender-ID\n carrier delivers a dynamic alphanumeric sender ID unchanged (about\n 110 countries, see core/hailhq/core/sender_id_corridors.py) use\n your organization's sender ID\n (or ``HAIL``, the platform default); Australia (+61) may use\n ``HAIL`` instead; every other destination,\n including the US and Canada,\n uses your organization's oldest active SMS-capable number and the\n API returns 422 when there is none.\n SMS never uses the shared voice-call pool.\n\n ``recipient_consent`` is required and must be the JSON boolean\n ``true`` (strings and numbers are refused): attest that you (the caller\n triggering this request) have obtained the lawful consent needed\n to text this recipient. The API rejects the request (422) if\n this is not ``true`` — Hail does not verify consent for you. Set\n ``message_type=\"marketing\"`` for promotional texts (this\n additionally requires a non-empty ``consent_source``) — leave as\n the default ``\"informational\"`` for transactional/service texts.\n\n ``idempotency_key`` defaults to a fresh UUID and is returned in\n the response under ``idempotency_key`` — pass the same value on\n a retry to replay rather than re-send.\n\n Example:\n send_sms(to=\"+14155551234\", body=\"Your order shipped!\",\n recipient_consent=True)\n\n Returns the ``SmsResponse`` dict (id, status, from_e164,\n to_e164, segment_count, ...). On failure returns\n ``{\"error\": \"<message>\"}`` instead.\n ",
"inputSchema": {
"properties": {
"body": {
"title": "Body",
"type": "string"
},
"consent_obtained_at": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Consent Obtained At"
},
"consent_source": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Consent Source"
},
"from_": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "From"
},
"idempotency_key": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Idempotency Key"
},
"message_type": {
"default": "informational",
"title": "Message Type",
"type": "string"
},
"metadata": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"title": "Metadata"
},
"recipient_consent": {
"title": "Recipient Consent",
"type": "boolean"
},
"to": {
"title": "To",
"type": "string"
}
},
"required": [
"to",
"body",
"recipient_consent"
],
"title": "send_sms_toolArguments",
"type": "object"
},
"name": "send_sms"
},
{
"description": "Send the user's personal or company details and identity documents\n to the carrier for verification. IRREVERSIBLE: once submitted it\n cannot be cancelled, and the carrier keeps the documents.\n\n ``attest_authorized``: pass true only if the person or company named\n in these details and documents asked for this submission and gave you\n the details and files in this conversation. Never invent, guess, infer\n or reuse personal data or documents. Anything else: pass false (the\n call then sends nothing).\n\n ``purpose`` is ``number`` (default) or ``sms``. ``purpose='sms'``\n registers the organization to text US numbers from a US number\n (country_code US, number_type local). The carrier fee is debited from\n credit at filing. Pass the same ``purpose`` to\n get_verification_requirements. For ``purpose='number'``, pass the\n same ``quote_id`` you passed to get_verification_requirements. For\n ``purpose='sms'``, pass the same ``number_id`` if you passed one.\n\n ``subject_type`` is required: ``person`` (the user) or ``business``\n (the user's company), the same one passed to\n get_verification_requirements. Never guess it: ask the user \"you or\n your company?\".\n\n The verification is named after the person or company in the form\n (for example the business name, or the first and last name); it\n comes back as ``subject_name``. The organization may hold several\n verifications of one subject type (several people or companies).\n\n Steps: 1) call get_verification_requirements. 2) Ask the user for each\n field and file. 3) Call this once. Do not submit twice: a 409 means\n one with the same subject type and name already exists (see\n list_verifications).\n\n Arguments, as listed by get_verification_requirements:\n ``fields`` is ``{field name: value}``. ``address`` is\n ``{\"customer_name\", \"street\", \"city\", \"region\", \"postal_code\",\n \"country_code\"}`` (all required when ``address_required``).\n ``documents`` is ``{slot: {\"option\": <option key>, \"fields\": {...}}}``\n for each slot with ``needs_input``. ``files`` is a list of\n ``{\"slot\", \"content_type\", \"content_base64\"}``, one per slot that\n needs a file; ``slot`` must be a key in ``documents``\n and match letters, digits, - and _ (at most 64).\n ``content_type`` is ``image/jpeg``, ``image/png`` or\n ``application/pdf``. Limits: 10 MiB per file, 29 MiB in total, at\n most 20 files and 20 fields. Compress images to JPG or PNG under about\n 7 MiB so the call stays small.\n\n Returns the verification: ``id``, ``state`` and ``subject_name``.\n Review takes business days and nothing notifies you: poll\n get_verification. If the carrier finds a problem you get\n ``{\"error\", \"problems\": [{\"loc\", \"msg\"}]}`` and nothing is stored:\n ask the user to fix it, then submit again. After a timeout call\n list_verifications before trying again.",
"inputSchema": {
"properties": {
"address": {
"anyOf": [
{
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"title": "Address"
},
"attest_authorized": {
"title": "Attest Authorized",
"type": "boolean"
},
"country_code": {
"title": "Country Code",
"type": "string"
},
"documents": {
"title": "Documents",
"type": "object"
},
"fields": {
"title": "Fields",
"type": "object"
},
"files": {
"items": {
"type": "object"
},
"title": "Files",
"type": "array"
},
"number_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Number Id"
},
"number_type": {
"title": "Number Type",
"type": "string"
},
"purpose": {
"default": "number",
"title": "Purpose",
"type": "string"
},
"quote_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Quote Id"
},
"subject_type": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Subject Type"
}
},
"required": [
"country_code",
"number_type",
"fields",
"documents",
"files",
"attest_authorized"
],
"title": "submit_verification_toolArguments",
"type": "object"
},
"name": "submit_verification"
},
{
"description": "Change a saved agent. Safe to repeat: the same call gives the same\n result.\n\n Only the fields you pass change; a field left out keeps its value.\n ``status`` is ``\"live\"`` or ``\"paused\"``; a paused agent does not\n answer. ``language`` is a lowercase ISO 639-1 code. Pass ``language``\n or ``voice_id`` alone: the other value is kept. ``max_duration_seconds``\n and ``handover_max_duration_seconds`` are 60..3600.\n ``handover_contacts`` replaces the whole list: ``[{contact_id, note}]``\n (max 10) as in ``create_agent``; ``[]`` removes every contact.\n\n To remove a value, list its name in ``clear_fields``. Allowed:\n ``first_message`` (the agent waits for the caller),\n ``ai_disclosure_line`` (back to the workspace line), ``tools``\n (back to all tools), ``max_duration_seconds`` (back to the workspace\n limit), ``handover_max_duration_seconds`` (back to 1800). Any other\n name is refused. Do not set and clear one field.\n\n Live calls keep the settings they started with. The change applies\n to the next call or text.\n\n Returns the updated ``AgentResponse`` as a dict, or\n ``{\"error\": \"<message>\"}``.",
"inputSchema": {
"properties": {
"agent_id": {
"title": "Agent Id",
"type": "string"
},
"ai_disclosure": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"title": "Ai Disclosure"
},
"ai_disclosure_line": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Ai Disclosure Line"
},
"clear_fields": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Clear Fields"
},
"first_message": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "First Message"
},
"handover_contacts": {
"anyOf": [
{
"items": {
"additionalProperties": true,
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Handover Contacts"
},
"handover_max_duration_seconds": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"title": "Handover Max Duration Seconds"
},
"language": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Language"
},
"max_duration_seconds": {
"anyOf": [
{
"type": "integer"
},
{
"type": "null"
}
],
"default": null,
"title": "Max Duration Seconds"
},
"name": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Name"
},
"recording_enabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"title": "Recording Enabled"
},
"sms_enabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"title": "Sms Enabled"
},
"status": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Status"
},
"system_prompt": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "System Prompt"
},
"tools": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"title": "Tools"
},
"voice_enabled": {
"anyOf": [
{
"type": "boolean"
},
{
"type": "null"
}
],
"default": null,
"title": "Voice Enabled"
},
"voice_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Voice Id"
}
},
"required": [
"agent_id"
],
"title": "update_agent_toolArguments",
"type": "object"
},
"name": "update_agent"
},
{
"description": "Upload a file to attach to a future outbound email.\n\n ``content_base64`` is the file's raw bytes, base64-encoded.\n Returns ``{\"id\": ..., \"filename\": ..., \"content_type\": ...,\n \"size_bytes\": ...}`` — pass ``id`` in ``send_email``'s\n ``attachment_ids`` list. The id is reusable across many sends\n and expires in 24h if never used. Files over 25MB (combined\n with the message body and any other attachments, per send) are\n rejected — host large files externally and link to them in the\n body instead.\n ",
"inputSchema": {
"properties": {
"content_base64": {
"title": "Content Base64",
"type": "string"
},
"content_type": {
"title": "Content Type",
"type": "string"
},
"filename": {
"title": "Filename",
"type": "string"
}
},
"required": [
"content_base64",
"filename",
"content_type"
],
"title": "upload_email_attachment_toolArguments",
"type": "object"
},
"name": "upload_email_attachment"
},
{
"description": "Identify the human whose credentials this session runs under.\n\n Use it to sign or route mail as that person: pass their address as\n ``send_email(reply_to=...)`` so replies reach them rather than the\n sending domain, which is often an unattended ``noreply@``.\n\n Returns ``{\"auth_kind\", \"organization_id\", \"user_id\", \"email\",\n \"name\"}``. ``user_id``/``email``/``name`` are null when the server\n runs on a shared operator key (``auth_kind=\"shared\"``) — there is no\n human behind it, so send without a ``reply_to``.\n\n Example:\n whoami()\n ",
"inputSchema": {
"properties": {},
"title": "whoami_toolArguments",
"type": "object"
},
"name": "whoami"
}
]
}What the leaf hashes
{"body":"aa604e5f8c71ee0ffff374cde1950b1c6811e3d3ff9323cd76f3d75ef87e1b6b","id":221994,"kind":"mcp.tools","method":"probe.mcp/1","observed_us":1793027752626515,"prev":null,"subject":"b028a026-1a23-4a6c-99ca-321c7d49e6a7","subject_kind":"endpoint"}Proof
Fold these into the leaf in order, hashing each sorted pair with keccak256, to reach the root of (296695 rows).
- 0x1dc764a49122746c4bb147ac938e7b648fab2e79b9ab1b475fd0d15248d8b1ab
- 0xa3d47bdeef2f86694ebeecf0451e7a1a00005cace1cd11bff5c87bc02ec79e81
- 0x87eb99b4749ebe2f700609245698217ba68b642400472e1474d6adf3e4d9f9db
- 0x40535dd4c52e9e2ae0b857f21d3bdfe9590151f5b66cf3776e466bdd24904bd3
- 0x830a262e808634a61fd99f26c1864f8ae9223aeab0e018f961b140514742061d
- 0x3c4f4751625001f48c1877990cf349d716694df7307016d2639b7897a9a0f173
- 0x91af5fae7e842077a75c117587ef6145d1d45fe50c228a2394c66ac0be50fef6
- 0x351963d5601e6f468dbfcd11606ef3ef2f6a02cd7eb3911628bfb5a86d56c499
- 0xe85b2c05abdf635cb7cb0997b20ea577db07b2f799012dcfafbf93151dfe1bb0
- 0x667c9edc7131bd388a1c1a039f7fddf0a4fbc13b4f1e7297601127a46cdc7292
- 0xf0f77ed1d484625ddaf2f6e58baa64c9dec47a22b1a6beb91219a17bddac7dfa
- 0x8e3531af6dd328ff9a174dd19751228f67fa0773c68a89edf859dffbd7b098c3
- 0x42b0d9d5b53af5b0fe99cee10ce29fe484db78b92ce49f97190cc6d098651d47
- 0x9dfd312f3d4937f7cb144cee3987e8516f222932b4ea286605236d73cfa1ed16
- 0xc3a6a3ae44ec0da1848b16d40c45d7aac97a67352eea8df18e0f26edabea8a2f
- 0x06d71bfa7d8b41d62c8836efc2c601e1cc3772d9453596b2ae4950887f3310ca
- 0x8c756542e482157df0c36464d75c937b845e4423dbbcfad08f0cf8d5a16b9a13
- 0x20704208aa6923d6b2450ef74ede87a6e0265d46447926038fe9506e9e154b2d
- 0x9b9ce5b74b9bab5121612ef6b64920a8d453cbf311096a423eef9702747e868a
- Day root
- 0xf43e8dcc3d451a4788b26be904c55e8c4c0fa75c9666796c51db38d37e2c0df8
- Chained
- 0x702d5ac508f08b0d1accf36af25822d8e5660367494edd17a642fb61e158a57a
- Published
- On Arc Testnet, block 69251483: 0x495162f8936e19f0039918feebe6c307f327cd6bf8c37219620b1970041fa9ae