Connecting an AI client to telmini

What it does

The telmini MCP server lets an AI assistant work on your telmini workspace: list and read conversations, reply to visitors, resolve threads, add website or question-and-answer knowledge, adjust the AI agent and keyword replies, manage FAQ and read analytics. Every tool acts only on the workspace that owns the key.

1. Create an MCP key

Sign in to your telmini workspace, open MCP in the sidebar and go to the Access & keys tab. Workspace owners and admins can create a key on every plan. The full key is shown once, so store it like a password. You can revoke it on the same page at any time.

2. Add the server to your AI client

The server speaks streamable HTTP at:

https://cs.askais.com/api/mcp/mcp

Claude Code:

claude mcp add --transport http askais https://cs.askais.com/api/mcp/mcp --header "Authorization: Bearer YOUR_MCP_KEY"

Cursor, in ~/.cursor/mcp.json:

{
  "mcpServers": {
    "askais": {
      "url": "https://cs.askais.com/api/mcp/mcp",
      "headers": { "Authorization": "Bearer YOUR_MCP_KEY" }
    }
  }
}

Gemini CLI, in settings.json (streamable HTTP uses httpUrl):

{
  "mcpServers": {
    "askais": {
      "httpUrl": "https://cs.askais.com/api/mcp/mcp",
      "headers": { "Authorization": "Bearer YOUR_MCP_KEY" }
    }
  }
}

Not available yet

Signing in to telmini from a ChatGPT connector is not available yet. Create your workspace on the telmini website first, then connect with a key as shown above.

Contact telmini support

Keep credentials private

Put the key only in your client configuration. Do not send passwords, payment details, API keys or one-time codes in an AI chat. Review the workspace and the requested action before your assistant replies to a visitor.