Skip to main content
Glama

Messaging: Start chat

messaging_start_chat
Destructive

Start a new personal WhatsApp, Instagram, Facebook or Telegram one-to-one/group chat (Facebook Pages cannot start chats) by sending its first message. Resolve users_ids first. A string starts 1-to-1; an array requests a group. LinkedIn requires the inbox-specific start tool.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "provider": "<provider>",
      "users_ids": "<users_ids>",
      "text": "<text>"
    },
    "name": "messaging_start_chat"
  }
}

Input Schema

NameRequiredDescriptionDefault
nameNo
textYes
providerYes
specificsNo
users_idsYesOne exact provider user ID for a direct chat, or exact provider user IDs for a group; resolve every recipient first.
account_idNoOptional Nilyo connection ID (unipile_account_id from list_connected_accounts). Omit when the user has one account for this provider. When several exist, Nilyo never guesses: list them (display name, identifier, provider user ID), choose the one the user named or ask, and pass its ID here.
attachmentsNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedInput schema / properties / provider / enum
      Previous value: -[
      -  "whatsapp",
      -  "instagram",
      -  "telegram"
      -]New value: +[
      +  "whatsapp",
      +  "instagram",
      +  "facebook",
      +  "telegram"
      +]
  2. Changed1 schema field changed
    • changedInput schema / properties / users_ids / anyOf
      Previous value: -[
      -  {
      -    "description": "Exact provider user ID returned by the relevant profile/contact resolver; never a display name or profile URL. For LinkedIn, use the stable system ID. LinkedIn system user ID: Stable LinkedIn/Unipile user id, commonly ACo... for Classic profiles. Obtain with: linkedin_get_profile_from_url(profile_url) -> result.id; linkedin_get_profile(user_id_or_public_identifier) -> result.id; linkedin_search_people(...) -> selected result.id Never pass: full linkedin.com/in/... URL, person name, company ID.",
      -    "minLength": 1,
      -    "type": "string"
      -  },
      -  {
      -    "items": {
      -      "description": "Exact provider user ID returned by the relevant profile/contact resolver; never a display name or profile URL. For LinkedIn, use the stable system ID. LinkedIn system user ID: Stable LinkedIn/Unipile user id, commonly ACo... for Classic profiles. Obtain with: linkedin_get_profile_from_url(profile_url) -> result.id; linkedin_get_profile(user_id_or_public_identifier) -> result.id; linkedin_search_people(...) -> selected result.id Never pass: full linkedin.com/in/... URL, person name, company ID.",
      -      "minLength": 1,
      -      "type": "string"
      -    },
      -    "minItems": 1,
      -    "type": "array"
      -  }
      -]New value: +[
      +  {
      +    "description": "Exact provider user ID returned by the relevant profile/contact resolver; never a display name or profile URL. For LinkedIn, use the stable system ID. LinkedIn system user ID: Stable LinkedIn user id, commonly ACo... for Classic profiles. Obtain with: linkedin_get_profile_from_url(profile_url) -> result.id; linkedin_get_profile(user_id_or_public_identifier) -> result.id; linkedin_search_people(...) -> selected result.id Never pass: full linkedin.com/in/... URL, person name, company ID.",
      +    "minLength": 1,
      +    "type": "string"
      +  },
      +  {
      +    "items": {
      +      "description": "Exact provider user ID returned by the relevant profile/contact resolver; never a display name or profile URL. For LinkedIn, use the stable system ID. LinkedIn system user ID: Stable LinkedIn user id, commonly ACo... for Classic profiles. Obtain with: linkedin_get_profile_from_url(profile_url) -> result.id; linkedin_get_profile(user_id_or_public_identifier) -> result.id; linkedin_search_people(...) -> selected result.id Never pass: full linkedin.com/in/... URL, person name, company ID.",
      +      "minLength": 1,
      +      "type": "string"
      +    },
      +    "minItems": 1,
      +    "type": "array"
      +  }
      +]
  3. First observed

TDQS

A4/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare destructive=true, openWorld=true and idempotent=false; the description adds the non-obvious fact that starting a chat actually delivers a message, and that Facebook Pages cannot start chats. It does not warn about duplicate-chat risk or how attachments affect the send, but covers the main behavioral surprises.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four short sentences, each carrying a distinct constraint (scope, prerequisite, mode rule, exclusion), front-loaded with the core action. No filler or restatement of the tool name.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a mutation tool with nested objects, 7 parameters and no output schema, the description omits any mention of media/attachment support and of choosing an account when several exist. The critical chat-creation semantics are complete, but an agent would not learn from the description that it can send files or voice notes.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters2/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is only 29%, so the description must compensate, and it does not: `name`, `specifics`, `attachments` (a nested object with filename/content/content_type/send_mode/voice_note) and `text` are never explained. Its users_ids string-vs-array explanation largely duplicates the schema's own description of that parameter, so value added beyond the schema is thin.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb (start), resource (one-to-one/group chat) and enumerates the providers it covers, plus explicitly excludes Facebook Pages and routes LinkedIn elsewhere. An agent can distinguish this from messaging_send_message or linkedin_start_conversation without opening a schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives a clear prerequisite (resolve users_ids first) and a crisp rule for the two modes (string = 1-to-1, array = group), plus an explicit exclusion with an alternative for LinkedIn. It never states the inverse condition — use messaging_send_message when a chat already exists — so the boundary with that sibling is only implied by 'by sending its first message'.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.