Integrations Guide

Formul8 provides several integration points for connecting its capabilities to your existing tools and workflows. This guide covers MCP tools for AI code editors, the embeddable knowledge widget, API integration patterns, and third-party connections.

MCP Tools for Cursor and Claude Desktop

Formul8 provides a Model Context Protocol (MCP) server that exposes 65+ METRC cannabis tracking tools to AI code editors like Cursor and Claude Desktop. This allows you to query and interact with METRC data directly from your development environment.

What Is MCP?

The Model Context Protocol is a standard for connecting AI assistants to external tools and data sources. When you connect Formul8's MCP server to Cursor or Claude Desktop, the AI assistant gains the ability to call METRC API operations on your behalf.

Available Tools (65+)

The MCP server provides tools across all major METRC API categories:

Analysis Skills

Beyond raw METRC API access, the MCP server includes 12 analysis skills that provide higher-level intelligence:

Connecting to Cursor

1. Open Cursor settings
2. Navigate to the MCP servers configuration
3. Add the Formul8 MCP server:

{
  "mcpServers": {
    "formul8-metrc": {
      "command": "npx",
      "args": ["-y", "formul8-metrc-mcp"],
      "env": {
        "METRC_VENDOR_API_KEY": "your-vendor-key",
        "METRC_USER_API_KEY": "your-user-key"
      }
    }
  }
}

4. Restart Cursor to activate the MCP connection
5. The AI assistant can now call METRC tools when you ask questions about your cannabis data

Connecting to Claude Desktop

1. Open Claude Desktop settings
2. Navigate to MCP server configuration
3. Add the server configuration (same format as Cursor above)
4. Restart Claude Desktop

Example MCP Interactions

Once connected, you can ask your AI assistant questions like:

• "List all active packages in my facility"
• "Are there any packages that need attention?"
• "Run an audit-ready check on my facility"
• "Show me the FIFO analysis for my flower inventory"
• "What is the yield analysis for my last 5 harvests?"

The AI assistant calls the appropriate MCP tools behind the scenes and presents the results in a readable format.

Vercel Edge API

The MCP server is also available as an HTTP API for custom integrations:

Base URL: https://metrc-mcp.vercel.app

Embeddable Knowledge Widget

The Future4200 knowledge widget (formul8future.js) can be embedded in your own website or application to provide AI-powered answers from the Future4200 knowledge base.

What It Does

The widget adds a chat bubble to your website. When a user clicks it, a chat panel opens where they can ask cannabis-related questions. Answers are sourced from 7,061 Future4200 forum threads with citations.

Embedding the Widget

Add the following script tag to your HTML:

<script src="https://future.formul8.ai/formul8future.js"></script>

The widget initializes automatically and places a chat bubble in the bottom-right corner of the page.

Customization

The widget supports configuration options passed as data attributes or through a JavaScript initialization call:

<script
  src="https://future.formul8.ai/formul8future.js"
  data-position="bottom-right"
  data-theme="dark"
></script>

Widget API

The widget communicates with the Future4200 knowledge API:

API Integration Patterns

OpenAI-Compatible Integration

Formul8's API follows the OpenAI Chat Completions format, making it compatible with any tool or library that supports the OpenAI API. See the API Reference for full details.

Drop-in replacement pattern:

If your application already uses the OpenAI API, you can switch to Formul8 by changing the base URL and API key:

# Before: OpenAI
client = OpenAI(api_key="sk-...")

# After: Formul8
client = OpenAI(
    api_key="YOUR_FORMUL8_API_KEY",
    base_url="https://chat.formul8.ai/v1"
)

All other code — message formatting, streaming, response parsing — remains the same.

Webhook Integration

The Dashboard (users.formul8.ai) processes Stripe webhooks for subscription events. If you need to integrate subscription changes into your own systems, you can configure Stripe webhooks in your Stripe dashboard to forward events to your own endpoints alongside Formul8's.

Key Stripe events handled:

• checkout.session.completed — New subscription created
• customer.subscription.updated — Plan upgrade/downgrade
• customer.subscription.deleted — Subscription cancelled
• invoice.payment_succeeded — Payment processed
• invoice.payment_failed — Payment failed

Health Monitoring

All Formul8 services expose health endpoints for monitoring:

You can integrate these into your own monitoring systems (Datadog, PagerDuty, etc.) to track Formul8 service availability.

Integration with Cannabis Systems

Metrc Track-and-Trace

Formul8 integrates with Metrc through two pathways:

1. Metrc Dashboard (metrc.formul8.ai) — Full UI dashboard with AI Copilot for interactive use
2. MCP Tools — 65+ tools for programmatic access via Cursor, Claude Desktop, or the HTTP API

Both pathways require Metrc vendor and user API keys, configured at the account level.

POS Systems

The Command Center supports CSV import from common cannabis POS systems:

• Dutchie
• Treez
• Flowhub

Export sales data from your POS system as CSV and import it into the Command Center for cross-facility analytics.

ERP and Accounting

Financial data can be imported into the Command Center via CSV for integration with:

• QuickBooks
• Sage
• Custom ERP systems

Coming Soon

Future4200 Knowledge API

The Future4200 knowledge base is available through a standalone API for custom integrations.

Base URL: https://future.formul8.ai

Search Threads

curl https://future.formul8.ai/api/threads?q=winterization+ethanol

Returns matching threads from the 7,061-thread corpus with relevance scores.

Chat with Context

curl -X POST https://future.formul8.ai/api/chat \
  -H "Content-Type: application/json" \
  -d '{
    "message": "What is the best winterization technique for crude ethanol extract?"
  }'

Returns an AI-generated answer with citations from relevant Future4200 threads.

OpenAI-Compatible

curl -X POST https://future.formul8.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "future4200",
    "messages": [
      {"role": "user", "content": "How do I troubleshoot a vacuum oven?"}
    ]
  }'

Rate Limits for Integrations

Next Steps

• API Reference — Full endpoint documentation with code examples
• Metrc Dashboard Guide — Interactive Metrc access
• Agent Catalog — Understand what you can query through the API

Last updated: 2026-02-18