🧠 RAGler

The human-in-the-loop RAG knowledge platform

Docs · API Reference · Report Bug · Request Feature

---

✨ What is RAGler?

RAGler is an open-source RAG knowledge operations platform that gives you full control over your retrieval-augmented generation pipeline — from ingestion to publishing.

• 📥 Ingest knowledge from web pages, file uploads, or manual text
• 🔍 Review & edit chunks in a draft session before they go live
• 🚀 Publish validated chunks to Qdrant vector collections
• 🤖 AI Agent for chat, collection cleaning, and chunk generation
• 🔌 MCP Server built-in — IDE integration (Cursor, VS Code, etc.) via /mcp endpoint

🎬 Demo

🏗️ Architecture

┌──────────────┐     ┌──────────────┐     ┌──────────────┐
│   Frontend   │────▶│   Backend    │────▶│    Qdrant    │
│  (Next.js)   │     │  (NestJS)    │     │  (vectors)   │
└──────────────┘     └──────┬───────┘     └──────────────┘
                            │
                     ┌──────┴───────┐
                     │    Redis     │
                     │  (sessions)  │
                     └──────────────┘

🚀 Quick Start

Prerequisites

Requirement	Version
Node.js	20+
pnpm	9+
Docker	latest
OpenAI Key	—

1. Clone & start infrastructure

git clone https://github.com/scrobot/ragler.git
cd ragler
docker compose up -d redis qdrant

2. Start backend

cd backend
pnpm install
cp .env.example .env
# set OPENAI_API_KEY in .env
pnpm start:dev

3. Verify

curl http://localhost:3000/api/health/liveness
# → {"status":"ok"}

4. Start frontend

cd frontend
pnpm install
pnpm dev
# → http://localhost:3000

5. (Optional) Verify MCP endpoint

curl -X POST http://localhost:3000/mcp \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","method":"tools/list","id":1}'

🐳 Docker Deployment

The fastest way to run RAGler — all services in one command.

Using pre-built images (recommended)

git clone https://github.com/scrobot/ragler.git
cd ragler

# Set your OpenAI key
echo "OPENAI_API_KEY=sk-your-key-here" > .env

# Start everything
docker compose up -d

This pulls pre-built images from ghcr.io/scrobot/ragler and starts:

Service	URL	Description
Frontend	localhost:3000	RAGler UI
Backend API	localhost:3010	API + Swagger
Qdrant	localhost:6333	Vector DB dashboard
Redis	localhost:6379	Session store

Build from source

docker compose up -d --build

Debug tools (RedisInsight)

docker compose --profile debug up -d
# → RedisInsight at http://localhost:5540

Stop & clean up

docker compose down           # stop containers
docker compose down -v        # stop + delete volumes

⚙️ Configuration

Backend config lives in backend/.env. Key variables:

Variable	Required	Default	Description
OPENAI_API_KEY	✅	—	OpenAI API key
REDIS_HOST	✅	—	Redis hostname
REDIS_PORT	—	6379	Redis port
QDRANT_URL	✅	—	Qdrant connection URL
PORT	—	3000	Backend port
SESSION_TTL	—	86400	Draft session TTL (seconds)
SQLITE_PATH	—	data/ragler.db	SQLite path for settings

🚩 Feature Flags

Toggle features via environment variables or the UI (Settings → Features):

Flag	Default	Controls
FEATURE_WEB_INGEST	true	Web URL ingestion
FEATURE_FILE_INGEST	true	File upload ingestion
FEATURE_AGENT	true	AI agent (chat, cleaning)

📖 Documentation

Resource	Link
📚 Full docs	scrobot.github.io/ragler
🔧 API Swagger	localhost:3000/api/docs
🏛️ Architecture	docs/architecture
🚀 Getting started	docs/getting-started

🧪 Testing

cd backend
pnpm test        # run all tests
pnpm lint        # lint check
pnpm typecheck   # type check

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the project
2. Create your feature branch (git checkout -b feat/amazing-feature)
3. Commit using conventional commits (git commit -m 'feat: add amazing feature')
4. Push to the branch (git push origin feat/amazing-feature)
5. Open a Pull Request

🐛 Troubleshooting

Problem	Solution
Readiness check fails	Check Redis/Qdrant: docker compose ps
401/403 errors	Ensure X-User-ID header is set
Ingest failures	Verify OPENAI_API_KEY and network connectivity

📄 License

Distributed under the ISC License.

---

_{Built with ❤️ by scrobot}