A lightweight, zero-dependency scheduler that lets command-line AI agents like Claude Code and Codex work in the background and ask you for decisions via cards only when needed.
rerun-build/rerun
Self-hostable platform for hiring AI coworkers that use your tools, run on schedules, and ask before sending, paying or deleting anything.
- Stars
- 790
- Forks
- 86
- Open issues
- 15
- Last commit
- 24 hours ago
What is it?
Rerun lets you describe work in plain language and have a planning agent called the Foreman assemble a team of AI coworkers. Each coworker has its own job, memory, skills and a browser-based computer that you can watch or take over. All coworkers appear on a shared board, and you can chat with them like colleagues, including file sharing and inline approvals.
Coworkers connect to external services through a catalog of 200+ integrations, any MCP server, and Composio, covering tools such as Gmail, Slack, Notion, HubSpot, Stripe and GitHub. They can run on schedules or webhook triggers, and they pause for human approval before sending messages, making payments or deleting data. Models can be supplied through API keys for Anthropic, OpenAI, Gemini or OpenRouter.
Self-hosting uses Docker, Node.js 22 and a Supabase project, with the app built on Next.js and an engine service that runs one per workspace. The same codebase powers the hosted Rerun Cloud, which adds a template library, managed computers and premium integrations. The project is licensed under AGPL-3.0.
Key features
- 01Foreman agent plans and assembles a team of coworkers from a plain-language request.
- 02Each coworker has its own memory, skills and browser-based computer you can watch or take over.
- 03Asks for approval before sending messages, making payments or deleting anything.
- 04Runs on schedules and webhook triggers without you present.
- 05Connects to 200+ integrations, any MCP server and Composio apps.
- 06Works with models from Anthropic, OpenAI, Gemini or OpenRouter.
Use cases
- Automating inbox triage, calendar management and daily briefings for an executive.
- Running a weekly KPI report from company data with charts.
- Drafting and scheduling social media posts for review before publishing.
- Finding sales prospects and drafting first outreach emails for approval.
Quick start
- 1
Clone the repository and enter the folder
git clone https://github.com/rerun-build/rerun.git && cd rerun - 2
Copy the example env file, then fill in the Supabase URL and keys and an OpenRouter key
cp .env.example .env - 3
Link your Supabase project and push the database schema with seed data
npx supabase link --project-ref <your-project-ref> && npx supabase db push --include-seed - 4
Build and start the app with Docker, then open http://localhost:3000
docker compose up -d --build
Things to know
- Self-hosting needs Docker, Node.js 22 and a Supabase project, plus API keys for the models you use.
- Self-hosting is documented but not supported by maintainers; help is community-only via Discord.
- Licensed under AGPL-3.0, so hosting it for others requires publishing your changes under the same license.
In collections
Topics
FAQ
- What is rerun?
- Self-hostable platform for hiring AI coworkers that use your tools, run on schedules, and ask before sending, paying or deleting anything. Rerun lets you describe work in plain language and have a planning agent called the Foreman assemble a team of AI coworkers. Each coworker has its own job, memory, skills and a browser-based computer that you can watch or take over. All coworkers appear on a shared board, and you can chat with them like colleagues, including file sharing and inline approvals.
- Is rerun free to use? What license does it have?
- rerun is open source under the AGPL-3.0 license. Check the license terms before using it in commercial products.
- Is rerun still maintained?
- Actively maintained. The last commit to rerun was on Nov 6, 2026.
- What can I use rerun for?
- Automating inbox triage, calendar management and daily briefings for an executive. Running a weekly KPI report from company data with charts. Drafting and scheduling social media posts for review before publishing. Finding sales prospects and drafting first outreach emails for approval.
- How do I install rerun?
- Clone the repository and enter the folder: git clone https://github.com/rerun-build/rerun.git && cd rerun Copy the example env file, then fill in the Supabase URL and keys and an OpenRouter key: cp .env.example .env Link your Supabase project and push the database schema with seed data: npx supabase link --project-ref <your-project-ref> && npx supabase db push --include-seed Build and start the app with Docker, then open http://localhost:3000: docker compose up -d --build
- Where can I get rerun?
- The source code and installation instructions are on GitHub: https://github.com/rerun-build/rerun
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A self-improving AI agent that creates and refines skills from experience, runs on many model providers, and works through CLI and messaging apps.
Self-hosted workspace for creating named AI coworkers with their own instructions, memory, routines, and approval controls.
Self-hosted, MIT-licensed sales CRM with native AI agents and WhatsApp integration via WAHA or Meta Cloud API, built on Next.js and Supabase.
Open-source platform for tracking how brands appear in AI search answers across ChatGPT, Gemini, Claude, Perplexity and other engines.
A toolkit of AI prompts that interview you to plan a product, plus a step-by-step guide for coding agents to build and launch a Next.js, Supabase and Vercel app.