ResearchChain

ResearchChain is a monorepo for running continuous AI-assisted research with worker services and a WebUI scheduler.

Quick Start

1) Start backend services

Run all workers, database, and utilities with Docker:

docker-compose -f docker/docker-compose.yml up

2) Start WebUI frontend

The frontend runs separately:

1. cd webui/frontend/
2. npm install
3. npm run dev
4. Open http://localhost:3000/

Postgres Access (pgAdmin)

pgAdmin is included in Docker Compose.

1. Open http://localhost:8081/browser/
2. Add a new server named postgres
3. In Connection:
   • Hostname/address: postgres
   • Username: admin
   • Password: pass
4. Save

Notes

• Recommended Python version: 3.9
• Default model configs: core/models/configurations
• environment.yml is for Linux; separate env files are available for Windows and macOS (Apple Silicon)

Architecture

• Research Chain: worker services that process and continue research tasks
• Schedulers: apps that create and manage tasks (including WebUI)

Diagrams

WebUI Flow

Database Schema

---

THIS SOFTWARE IS INTENDED FOR EDUCATIONAL AND RESEARCH PURPOSES ONLY.
WE ARE NOT RESPONSIBLE FOR ANY ILLICIT USES OF THIS SOFTWARE.