Cb Insights MCP Server

What is Cb Insights MCP Server about?

The server exposes a set of HTTP endpoints that let developers send messages to the ChatCBI large language model and receive structured responses, including the conversation ID, related content, source citations, and suggested follow‑up prompts.

How to use Cb Insights MCP Server?

Install dependencies – the project uses the Python package manager uv. After cloning the repository, run uv sync (or the equivalent uv install).
Configure environment variables – copy the .env.example file to .env and set:
- CBI_CLIENT_ID and CBI_CLIENT_SECRET for API authentication.
- Optional: CBI_MCP_PORT (default 8000) and CBI_MCP_TIMEOUT.
Start the server – execute the command generated by mcp install server.py or run manually:
```
uv --directory /path/to/cloned/cbi-mcp-server run server.py
```
Call the API – send POST requests to /chatcbi with message and optional chatID fields. The response contains chatID, message, RelatedContent, Sources, and Suggestions.

Key features of Cb Insights MCP Server

Session continuity – optionally provide a chatID to keep context across multiple calls.
Rich response payload – includes related content, source references, and suggested prompts.
Configurable timeout and port via environment variables.
Simple installation with the lightweight uv manager.
Built‑in debugging using the Model Context Protocol inspector (mcp dev server.py).

Use cases of Cb Insights MCP Server

Integrating ChatCBI into custom AI agents or chat‑bots.
Building research assistants that need citation‑backed answers.
Creating interactive dashboards that surface related market intelligence.
Automating follow‑up question generation for analysts.

FAQ from the Cb Insights MCP Server

Do I need a CB Insights account? Yes, you must have a valid client ID and secret to authenticate with the CB Insights API.
Can I change the default port? Set CBI_MCP_PORT in the .env file to any available port.
What happens if I omit chatID? The server creates a new ChatCBI session and returns its chatID in the response.
How do I debug the server? Use the Model Context Protocol inspector: mcp dev server.py.
Is there a timeout for requests? Yes, configurable via CBI_MCP_TIMEOUT (seconds).

The CBI MCP Server provides an interface for developers to interact with CB Insights ChatCBI LLM through AI Agents.

Tools

ChatCBI

Sends a message from an agent to our AI chatbot and returns the response
Input parameters:
- message:
- chatID: (optional) The unique id of an existing ChatCBI session. Used for continuity in a conversation. If not provided, a new ChatCBI session will be created
Returns object containing the following fields:
- chatID: Unique id of current ChatCBI session
- message: ChatCBI message generated in response to the message send in the input.
- RelatedContent: Content that is related to the content returned
- Sources: Supporting sources for the message content returned
- Suggestions Suggested prompts to further explore the subject matter
For more information, check the ChatCBI Docs

Setup

The CBI MCP Server uses uv to manage the project.

The default port is 8000, but can be modified by updating the CBI_MCP_PORT environment variable in the .env file.

The timeout for requests can also be modified via the CBI_MCP_TIMEOUT variable in the .env file.

Authentication

Documentation on how CB Insights APIs are authenticated can be found here

The server uses the CBI_CLIENT_ID and CBI_CLIENT_SECRET environment variables set in the .env file to authorize requests.

Usage

With Claude Desktop

Update the claude_desktop_config.json file using the following command:

mcp install server.py

This will add the following configuration:

{
  "mcpServers": {
    "cbi-mcp-server": {
      "command": "/path/to/.local/bin/uv",
      "args": [
        "--directory",
        "/path/to/cloned/cbi-mcp-server",
        "run",
        "server.py"
      ]
    }
  }
}

Debugging

The inspector can be used to test/debug your server.

mcp dev server.py

More info on using the inspector

Cb Insights MCP Server