MCP Integration

Connect VectorForgeAI to Claude, ChatGPT, and other MCP (Model Context Protocol) clients to search your team's internal documentation directly within your AI workflows.

NEW Build your own RAG-enabled document MCP for your team's docs

What is MCP?

Model Context Protocol (MCP) is an open standard that enables AI assistants to securely connect to external data sources and tools. VectorForgeAI provides an MCP server that allows you to search your documents directly from within AI assistants like Claude Code.

This integration provides a powerful alternative to services like ref.tools and Context7, specifically designed for teams who want to search their internal documentation alongside any public documentation they've added to their collections.

Key Features

Search across all your uploaded documents
Seamless integration with Claude, ChatGPT, and other MCP clients
Uses your existing document collections - no additional setup required
Efficient billing: only 1 search credit per MCP search_docs call
Secure authentication with your existing API tokens

Setup Instructions

For Claude Code

Add VectorForgeAI as an MCP server in Claude Code with a single command:

Terminal

claude mcp add --transport http "VectorForgeAI" \
  "https://api.vectorforgeai.com/mcp" \
  --header "Authorization: Bearer YOUR_API_TOKEN" \
  --header "Team-Token: YOUR_TEAM_TOKEN"

⚠️ Important

Replace YOUR_API_TOKEN with your actual VectorForgeAI API token. You can find your API token in your api tokens list.

Replace YOUR_TEAM_TOKEN with your Team Token. You can find it in Team Settings inside the dashboard.

For Other MCP Clients

If you're using a different MCP client, configure it with the following endpoint details:

Transport	HTTP (JSON-RPC)
Endpoint	https://api.vectorforgeai.com/mcp

The MCP endpoint uses JSON-RPC 2.0 format and handles all MCP methods including:

initialize - Initialize the MCP connection
tools/list - List available tools
tools/call - Execute a tool

Include the following headers with each request:

Authorization: Bearer YOUR_API_TOKEN
Team-Token: YOUR_TEAM_TOKEN

Available Tools

The VectorForgeAI MCP server currently provides the following tool:

search_docs

Search through your document collections for relevant content.

Usage Example

Once configured, you can use natural language in Claude Code to search your documents:

Claude Code Session

# CLAUDE.md
## Documentation Collections
- Authentication: 1234abcdef

# User
Search for our API authentication documentation

# Claude (using MCP)
I'll search your documentation for API authentication information...

[Searches through your VectorForgeAI documents in collection 1234abcdef]

Based on your documentation, here's what I found about API authentication:
- All API requests require a Bearer token in the Authorization header
- API tokens can be generated in your team settings
- Tokens should be kept secure and never exposed in client-side code
...

Billing

MCP integration uses your existing document search credits:

1 search credit per search_docs call
No additional charges for MCP protocol overhead
Same pricing as regular API searches

Security Considerations

🔒 Security Best Practices

Keep your API token secure and never share it publicly
Use separate API tokens for different environments (dev/prod)
Regularly rotate your API tokens
MCP requests are authenticated with your team's API token
All searches are scoped to your team's documents only

Troubleshooting

Common Issues

Authentication Failed

If you receive authentication errors, verify that:

Your API token is correctly formatted with "Bearer " prefix
The token hasn't expired or been revoked
You're using the correct team's API token

No Search Results

If searches return no results:

Ensure you have documents uploaded to your collections
If you have recently uploaded, wait up to 5 minutes for embedding to finish
Try broader search terms

Connection Errors

If the MCP client can't connect:

Verify the base URL is correct: https://api.vectorforgeai.com/mcp
Check your internet connection
Ensure your firewall allows HTTPS connections to our API

Next Steps

Upload Documents - Add more content to search through MCP
Manage Collections - Organize your documents for better search results
API Settings - Generate and manage your API tokens

API Documentation