> For the complete documentation index, see [llms.txt](https://docs.generationlab.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.generationlab.org/basics/connecting-your-verbai-mcp-server.md).
# Connecting Your VerbAI MCP Server
This guide explains how to connect your personal VerbAI MCP (Model Context Protocol) server to supported AI clients, giving you direct access to your behavioral data for analysis, querying, and programmatic workflows.
VerbAI exposes your personal data through an MCP server hosted on Snowflake. You can connect to it from multiple clients depending on your workflow:
* **Claude.ai** — Browser-based, no setup required (OAuth)
* **Claude Code / Claude Desktop** — Terminal or desktop app (Bearer token)
* **Gemini CLI** — Command-line interface with large context window
Each method requires credentials provisioned by VerbAI. Contact your VerbAI account manager to get set up.
***
### Option 1: Claude.ai (OAuth)
The simplest option. Claude.ai supports MCP connectors natively through OAuth — no local tooling required.
#### Steps
1. Go to [claude.ai](https://claude.ai/) → **Settings** → **MCP Connectors**
2. Click **Add a new connector** and enter the following:
* **Server URL:** Your VerbAI MCP server URL (provided by your account manager)
* **Client ID:** Your OAuth Client ID
* **Client Secret:** Your OAuth Client Secret
3. When prompted, log in with the Snowflake credentials provided to you.
* You will be asked to change your password on first login.
4. Once connected, you can query your VerbAI data directly in any Claude.ai conversation.
> **Note:** OAuth credentials and Snowflake login details are provisioned per-user by VerbAI. Do not share them.
***
### Option 2: Claude Code / Claude Desktop (Bearer Token)
For developers who prefer terminal workflows (Claude Code) or the Claude Desktop app, you can register the MCP server using a bearer token.
#### Prerequisites
* [Claude Code](https://docs.anthropic.com/en/docs/claude-code) or [Claude Desktop](https://claude.ai/download) installed
* Node.js (v18+) with `mcp-remote` installed globally:
```bash
npm install -g mcp-remote
```
#### Configuration
Add the following to your MCP configuration file:
* **Claude Code:** `~/.claude/mcp.json`
* **Claude Desktop:** `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) or `%APPDATA%\Claude\claude_desktop_config.json` (Windows)
```json
{
"mcpServers": {
"verb-ai-mcp": {
"command": "mcp-remote",
"args": [
"YOUR_VERBAI_MCP_SERVER_URL",
"--header",
"Authorization: Bearer YOUR_BEARER_TOKEN"
],
"env": {
"MCP_SERVER_REQUEST_TIMEOUT": "300000"
}
}
}
}
```
Replace:
* `YOUR_VERBAI_MCP_SERVER_URL` with the MCP server URL provided by your account manager
* `YOUR_BEARER_TOKEN` with your personal bearer token
> **Note:** If `mcp-remote` is not on your PATH, use the full path to the binary (e.g., the output of `which mcp-remote` or `where mcp-remote`).
> ⚠️ **Do not share your bearer token.** It grants full access to your personal data.
#### Verify
In Claude Code, run:
```bash
claude mcp list
```
You should see `verb-ai-mcp` listed and connected.
***
### Option 3: Gemini CLI
Gemini CLI offers a large context window and strong MCP support, making it a good fit for pulling and processing large volumes of data.
#### Prerequisites
Download and install the Gemini CLI:
#### Add Your VerbAI MCP Server
Run the following command in your terminal:
```bash
gemini mcp add --transport http verbai \
https://api.generationlab.org/mcp \
--header "Authorization: Bearer YOUR_MCP_API_KEY"
```
Replace `YOUR_MCP_API_KEY` with the API key provided by your account manager.
> ⚠️ **Do not share this token.** It grants full access to your personal data.
#### Verify
```bash
gemini mcp list
```
You should see your VerbAI MCP listed with a green checkmark and a status of **Connected**.
#### Start Querying
Launch interactive mode:
```bash
gemini
```
Then query your data. For example:
```
Using verbai MCP, pull search behavioral data about searches done in the past 24 hours
```
> **Tip:** Explicitly instruct Gemini to use the VerbAI MCP at the beginning of your prompt — otherwise it may respond without querying your data.
#### Approve Tool Execution
On first use, Gemini will ask whether it can run MCP tools. Use your arrow keys to select **"Yes, always allow\..."** and press Enter. This approval is required only once.
***
### Integration with n8n
You can also connect your VerbAI MCP server to [n8n](https://n8n.io/) for building automated pipelines, scheduled data pulls, and visual reports. Contact your account manager for integration guidance.
***
### Getting Access
All connection methods require credentials provisioned by VerbAI. To get started:
1. Contact your VerbAI account manager
2. You'll receive:
* Your MCP server URL
* A bearer token (for Claude Code, Claude Desktop, or Gemini CLI)
* OAuth credentials (Client ID + Secret) and Snowflake login (for Claude.ai)
3. Choose the connection method that fits your workflow and follow the steps above
***
### Troubleshooting
**Connection timeout:** Increase the timeout by setting `MCP_SERVER_REQUEST_TIMEOUT` to a higher value (in milliseconds). The default config above uses 300000 (5 minutes).
**Bearer token expired:** Tokens have an expiration date. Contact your account manager for a fresh token if you get authentication errors.
**OAuth login issues:** If your Snowflake password has expired or you're locked out, reach out to your account manager to reset credentials.
**Gemini not using MCP tools:** Make sure your prompt explicitly references the VerbAI MCP. Gemini defaults to its own knowledge unless directed to use external tools.
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://docs.generationlab.org/basics/connecting-your-verbai-mcp-server.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.