MCP: Overview
Verboo's Model Context Protocol (MCP) exposes 22 tools that let you create and manage assistants, configure triggers, add actions, test conversations end-to-end, and analyze history: all through natural language, without opening the dashboard.
It works with any MCP client that supports HTTP transport: Claude Desktop, Claude Code, Gemini CLI, Antigravity, and others.
Endpoint
https://api.verbeux.com.br/mcpAuthentication: Api-Key header with your API Key (see Authentication). The MCP uses the exact same key as the REST API: every operation runs on behalf of your company, with the same limits and permissions.
How it works
- You connect your MCP client to Verboo's endpoint (one-time setup).
- The 22 tools become available inside the client: the LLM sees them alongside its native tools.
- You ask in plain language ("create a sales assistant", "list triggers for agent 42") and the LLM decides which tools to call, in what order, and with what arguments.
- Results flow back to the LLM, which interprets them and replies: or chains more calls if needed.
Tools at a glance
| Category | Tools | What it's for |
|---|---|---|
| Assistants | 4 | List, create, update, and configure agents |
| Triggers | 5 | Define the functions the LLM can call during a conversation |
| Actions | 5 | Configure what each trigger executes (webhook, CRM, Feegow, etc.) |
| Sessions | 4 | Create test conversations, send messages, and inspect responses |
| History | 4 | List production messages and sessions, analyze volume and usage |
Next steps
- Client Setup: Claude Desktop, Claude Code, Gemini CLI, Antigravity
- Tools: Full reference of the 22 tools with parameters
- Example Conversations: 3 real end-to-end usage scenarios