Analysis updated 2026-05-18
Schedule a follow up research task so an AI assistant checks back on a question later and emails the answer.
Give Claude the ability to run scheduled web searches on your behalf.
Build automated reminders that combine AI research with email delivery.
Learn how to build an MCP server that exposes scheduling actions to an AI assistant.
| cpenniman12/alerts-mcp | 0xradioac7iv/tempfs | abboskhonov/hermium | |
|---|---|---|---|
| Stars | 0 | 0 | 0 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | hard | moderate | moderate |
| Complexity | 4/5 | 3/5 | 4/5 |
| Audience | developer | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires accounts and API keys for Anthropic, Tavily, Resend, and Supabase.
Alerts-MCP is a server that gives an AI assistant like Claude the ability to schedule research tasks and deliver the results to your inbox. Instead of waiting for you to check back on a question whose answer is not yet available, the assistant can register a reminder, and a background process will look into it on a schedule you choose and email you when it finds something. The setup involves two pieces running at the same time. The first is the MCP server, which runs locally on your computer and connects to your AI assistant, giving it four actions it can call: create an alert, list existing alerts, delete one, or run one immediately. The second is a worker process that runs continuously, checking a database every 30 seconds for alerts that are due and then executing a small AI-driven research loop using web search to investigate the topic. A concrete example from the README: you ask Claude about festivals in a city during a future trip. It says the schedule has not been announced yet. With this tool, it can schedule a follow-up in two weeks, and when the worker runs at that time, it searches the web, writes up what it found, and sends you an HTML email with the result. The project requires accounts with several external services. It uses the Anthropic API for the AI parts of the research loop and the alert parsing, Tavily for web search, Resend to send emails, and a Postgres database through Supabase to store scheduled alerts. The worker is designed to be deployed on a hosting service called Railway so it keeps running even when your computer is off, while the MCP server itself stays on your local machine. The project is early stage with no stars and a brief codebase. The README covers setup, the required environment variables, and how to register the server with Claude Desktop or Claude Code. No license information is mentioned in the README.
An MCP server that lets an AI assistant schedule follow-up research tasks and email you the results later.
Mainly TypeScript. The stack also includes TypeScript, MCP, Anthropic API.
Setup difficulty is rated hard, with roughly 1h+ to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.