sequence-mcp-server

alanwyatt3d/sequence-mcp-server

3.1

If you are the rightful owner of sequence-mcp-server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

This repository provides a minimal implementation of a Model Context Protocol (MCP) server for integrating Sequence.io accounts with ChatGPT, using FastAPI for deployment.

Sequence MCP Server Starter

This repository contains a minimal implementation of a Model Context Protocol (MCP) server for integrating Sequence.io accounts with ChatGPT. The server is built with FastAPI and can be deployed on a free tier at Render or any other hosting provider.

Features

  • Search (/mcp/search): Search for Sequence accounts by name or ID. Returns a list of results in the MCP content format.
  • Fetch (/mcp/fetch): Retrieve the full record for a Sequence account or a placeholder description for rules.
  • Remote amount (/remote/amount): Compute a transfer amount based on a configurable checking buffer, sweep percentage, and daily cap. Suitable for Sequence's “Query Remote API” action.
  • Trigger rule (/rules/{rule_id}/trigger) (optional): Trigger a Sequence rule via remote API. Protected by an admin token and rule-specific secrets.

Deployment

This repository includes a render.yaml file for quick deployment on Render. After uploading the repository to GitHub:

  1. Create a new Web Service on Render and connect your GitHub repository.
  2. Set the build command to pip install -r requirements.txt and the start command to uvicorn app:app --host 0.0.0.0 --port $PORT.
  3. Add the required environment variables in the Render dashboard:
    • SEQUENCE_ACCESS_TOKEN: Your Sequence access token.
    • SEQUENCE_RULE_SECRETS_JSON (optional): JSON map of rule IDs to API secrets.
    • MCP_ADMIN_TOKEN (optional): A secret token to secure rule triggering.
    • SWEEP_CHECKING_BUFFER: Minimum checking balance before sweeping excess.
    • SWEEP_PERCENT: Percentage of excess to sweep into savings/investing.
    • SWEEP_DAILY_CAP_CENTS: Maximum cents to sweep per day.
  4. Deploy the service. Once live, your MCP endpoints will be available under your Render domain (e.g. https://your-service.onrender.com).

Usage

Configure ChatGPT to use your MCP server by adding it as a connector in ChatGPT’s settings. Allow the search and fetch tools. For Sequence rule actions, point the rule’s Query Remote API URL to /remote/amount on your service, and optionally use the rule trigger endpoint with an admin token.

This starter kit is provided as-is and is intended to accelerate the setup of personal or small-scale Sequence integrations. Customize the code to suit your exact workflow needs.