Omnichannel MCP Server for DeepSeek

Connect your DeepSeek assistant to ChatbotX via MCP SSE. Manage customer conversations, CRM contacts, tags, and chatbot flows across your connected customer channels.

Read the Docs Start for $0

Please make sure there is always a human in the loop.

Connect ChatbotX MCP to DeepSeek

Paste the prompt command into your DeepSeek interface to connect automatically, or configure it manually using the settings block below.

Run in DeepSeek CLI / Prompt

> deepseek mcp add chatbotx --sse https://app.chatbotx.io/mcp/sse --header "x-workspace-token: YOUR_WORKSPACE_TOKEN"

or configure manually

Choose Client App

Select your preferred desktop or IDE environment for DeepSeek

DeepSeek Web is chat-only. To empower DeepSeek with MCP tools, connect via native MCP desktop clients like Cherry Studio or Chatbox:

Go to Settings > Model Provider > Select DeepSeek and enter your DeepSeek API key.
Navigate to MCP Servers > Click Add MCP Server.
Enter the ChatbotX connection parameters:

Server URL (Type: SSE)

https://app.chatbotx.io/mcp/sse

Header

x-workspace-token: YOUR_WORKSPACE_TOKEN

Configure Cline (VS Code Extension) to use DeepSeek V3/R1 models with ChatbotX MCP:

In Cline Settings, set API Provider to DeepSeek (or OpenAI-Compatible with api.deepseek.com).
Open the MCP Servers tab and add the server configuration:

{
  "mcpServers": {
    "chatbotx": {
      "url": "https://app.chatbotx.io/mcp/sse",
      "headers": {
        "x-workspace-token": "YOUR_WORKSPACE_TOKEN"
      }
    }
  }
}

DeepSeek MCP Integration

DeepSeek reads this page and wires up the MCP. You approve the connection.

It fetches this page, discovers the MCP endpoint, and asks you to confirm before connecting.

chatbotx.io/deepseek

# ChatbotX Model Context Protocol (MCP) Server for DeepSeek

Connect DeepSeek V3 and R1 reasoning models to ChatbotX via MCP SSE using Cherry Studio, Chatbox, or Cline. Leverage DeepSeek deep reasoning capabilities to resolve complex customer support cases, analyze policy edge cases, and execute precise CRM actions.

## Connection Setup for DeepSeek
- Protocol: Model Context Protocol (MCP) via Server-Sent Events (SSE)
- Endpoint: https://app.chatbotx.io/mcp/sse
- Authentication: Header "x-workspace-token: <YOUR_WORKSPACE_TOKEN>"
- Documentation: https://chatbotx.io/docs/mcp/introduction

### Configuration (Cherry Studio / Cline / Chatbox)
```json
{
"mcpServers": {
"chatbotx": {
"url": "https://app.chatbotx.io/mcp/sse",
"headers": {
"x-workspace-token": "YOUR_WORKSPACE_TOKEN"
}
}
}
}
```

## Available MCP Tools
- contacts_search: Retrieve contact profiles by email, international phone number, or workspace ID.
- contacts_tag: Execute logical tagging based on multi-step reasoning deductions.
- contacts_custom_field: Update custom fields (e.g., warranty_status, risk_score).
- flows_trigger: Trigger visual chatbot flows programmatically.
- conversations_list: Read multi-message conversation histories with rich metadata.
- messages_send: Draft and send rigorously validated responses across channels.
- broadcasts_list: Query broadcast histories and analyze campaign engagement data.
- channels_list: Inspect operational health of all 9 connected messaging channels.

## Deep Reasoning Workflows for DeepSeek
- "Evaluate dispute on order #4912 from contact '[email protected]'. Check conversation history against return policy and draft an optimal resolution."
- "Extract structured customer requirements from a complex 20-message chat on WhatsApp and update their CRM custom fields."
- "Analyze unassigned customer tickets in Messenger, group them by root cause, and tag contacts accordingly."

## Supported Messaging Channels
WhatsApp, Messenger, Instagram, TikTok, Telegram, Zalo OA, Webchat, SMTP, and API.

Prompt Examples for DeepSeek

Once connected, you can converse with your customer base and execute omnichannel actions in plain language.

Send Omnichannel Message

"Send a WhatsApp message to +123456789 saying their order #9821 has shipped with tracking link."

Query CRM & Contacts

"Find the contact for [email protected], check their conversation history, and tag them as 'VIP Lead'."

Trigger Visual Flows

"Trigger the Onboarding Drip Sequence for the newly registered lead on Instagram."

Audit Live Inbox

"List all unread conversations in Messenger and summarize what customers are asking today."

Autonomous Multi-Channel Execution

ChatbotX standardizes your messaging channels into clean MCP tool calls with safe guardrails.

Unified MCP Architecture

One SSE endpoint provides 380+ tools covering WhatsApp, Messenger, Zalo, Telegram, Email, and Webchat.

Enterprise Security

Workspace token authentication, granular tool whitelisting, and strict human-in-the-loop approvals for sensitive actions.

Real-time State Sync

Changes to contact fields, custom tags, or conversation status sync immediately across your ChatbotX dashboard and CRM.

Works with Every MCP Client

ChatbotX MCP works seamlessly with any AI agent runtime that supports Model Context Protocol.

Kimi K3

Moonshot AI Agent

Moonshot AI agent with 2M token context. Analyze complete conversation histories and trigger multi-step chatbot flows.

Explore Kimi K3 Integration

Muse

Creative & Support Agent

Specialized creative and workflow AI agent. Crafts high-converting customer responses and manages messaging sequences.

Explore Muse Integration

Supported Messaging & Customer Channels

Connect your channels in ChatbotX once, then automate customer conversations to any of them via your AI assistant.

Messenger

Instagram

TikTok

Zalo OA

Webchat

SMTP

API