Connect Tokens for Total Voice and ShopOS Accounts

Total Voice + AI Access

Connect FlatRate Pro profile, stats, estimates, diagnostics, and job status updates to ChatGPT, Claude, MCP servers, and more.

MCP is the preferred route. Every token is linked to your account, scoped to your live data, and revocable at any time. The public FlatRate Pro endpoint is read-first with limited safe writes: structured job logging plus technician job status updates.

💬 Open Custom GPT 🧩 MCP setup (preferred) 📄 Setup guide

Status Checking

Verifying Connect Token access…

Hang tight while we confirm your Total Voice or ShopOS API access for Connect Tokens.

ChatGPT Actions

Custom GPT Integration

Create a token below (label it "chatgpt").
In GPT Builder, go to Configure → Actions.
Click Import from URL and paste:
https://geargenius.ai/.well-known/frp-gpt-openapi.json
In Authentication, select API Key → Custom.
Header Name: X-User-Token
Paste your token as the Key.
Use it for profile, recent jobs, stats, structured job logging, job status updates, labor estimates, DTC lookup, and vehicle lookup.

Claude Projects

Anthropic Claude

Create a token (label it "claude").
Add API integration in Claude Project settings.
Use token in X-User-Token header.

MCP Server

Model Context Protocol (Preferred)

Create a token (label it "mcp").
Start with hosted MCP profile JSON, then apply agent policy.
Pass token via X-User-Token header.
Optional: run the reference local MCP bridge for controlled environments.

Technician profile: /.well-known/frp-mcp-config.json

Manager profile: /.well-known/shopos-org-mcp-config.json

Agent profiles: /mcp/agent-profiles.json

Bridge setup: /mcp-server-setup.html

Zapier / Make

No-Code Automation

Create a token (label it "zapier").
Use Webhooks by Zapier or HTTP module.
Add X-User-Token header.

Custom Apps

REST API

Use the same token with any HTTP client, script, or app for the public profile, lookup, and structured job logging tools.

curl -H "X-User-Token: TOKEN" \

https://geargenius.ai/api/gpt/user/stats

LangChain / AI Agents

Agent Frameworks

Create a token (label it "agent").
Configure API tool with base URL.
Inject token via request headers.

Token creation

Use one token per person and use case, treat tokens like passwords, and revoke immediately

Generate a Connect Token

MCP is the recommended default route. Tokens work with all platforms above. Label them by use case. Tokens are shown once, so store them securely and revoke them if exposed.

Label Expires in (days)

Your tokens

Active Connect Tokens

0 tokens

Label	Token ID	Expires	Created	Actions
No tokens yet.

Best practices

Use separate labels for ChatGPT, Zapier, internal bots, etc.
Rotate tokens annually or sooner if sharing with contractors.
Always revoke unused tokens to minimize blast radius.

Need help?

Review the Actions API guide and MCP setup guide. Public tools are read-first with two explicit safe writes: create job and update job status. Tokens are hashed and revocable at any time.