fagoon-upgrade 2.0.4

Fagoon AI Agents Workflow

Fagoon Upgrade

Self-hosted AI platform with workflow automation, agents, chat, and API generation.

Build AI workflows visually, publish them as REST APIs, deploy agents, and chat with LLMs — all running on your own machine.

---

Quick Start

Prerequisites

• Docker Desktop — Download here
• Python 3.10+ — Download here
• pipx (recommended) — Install with: pip install pipx

Install & Run

<<<<<<< HEAD
pipx install fagoon-upgrade
fagoon up                 # lite mode: no Redis, no Celery, zero env required
fagoon up --ollama        # add the local Gemma fallback LLM

Open http://localhost:8000 and complete the setup wizard (LLM key + features). Secrets are auto-generated; the database URL and config live in ~/.fagoon.

Install the CLI

pipx install fagoon-upgrade

feat/dual-mode

Start the platform

fagoon up

On first run, you'll see a setup wizard:

Welcome to Fagoon! Let's set up your AI platform.

How would you like to power your AI?

1. I have an API key (OpenAI, Gemini, Groq, etc.)
2. Use local Ollama model (free, runs on your machine)
3. Skip for now (configure later in the UI)

Choose [1/2/3]:

Pick your option and Fagoon starts automatically.

### Access the Platform

| Service  | URL                          |
|----------|------------------------------|
| Frontend | http://localhost:3000        |
| Backend  | http://localhost:8000        |

---

## Features

### Visual Workflow Builder
Drag-and-drop workflow canvas to build AI automations. Connect LLM nodes, filters, loops, and integrations visually.

### Workflow-as-API
Publish any workflow as a REST API endpoint with one click. Get an API key and endpoint URL — call it from any app, any language.

```bash
curl -X POST https://your-domain/api/v1/workflow-api/my-workflow/execute \
  -H "X-API-Key: wfapi_xxxxx" \
  -H "Content-Type: application/json" \
  -d '{"input": "Summarize this document"}'

AI Agents

Build and deploy conversational AI agents with memory, tool use, and multi-turn conversations.

Chat Interface

Chat with any LLM provider (OpenAI, Gemini, Groq, Anthropic, Ollama) through a unified interface.

Multi-Provider Support

Use any combination of LLM providers. Switch between them per workflow or per agent.

---

CLI Commands

fagoon up                    # Start the platform (lite mode)
fagoon up --full             # Start with Redis + Celery (multi-worker)
fagoon up --ollama           # Start with local Ollama LLM
fagoon down                  # Stop the platform
fagoon logs -f               # Follow live logs
fagoon status                # Check if running
fagoon update                # Pull latest version and restart
fagoon config set KEY=VALUE  # Set configuration
fagoon config show           # Show current config
fagoon db set-url URL        # Switch database

---

Configuration

Configuration is stored in ~/.fagoon/config.json. Set values via CLI or the web UI.

LLM Provider Keys

fagoon config set GEMINI_API_KEY=your-key-here
fagoon config set OPENAI_API_KEY=your-key-here
fagoon config set GROQ_API_KEY=your-key-here

Or configure through the web UI at http://localhost:3000 after login.

Using Local Ollama

Run with the Ollama profile to use free local models:

fagoon up --ollama

This starts an Ollama container alongside Fagoon. The Gemma 2B model is available as a fallback when no API keys are configured.

---

Deployment Modes

Lite Mode (Default)

Single-user, single-process. No Redis or Celery required. Perfect for personal use and small teams.

fagoon up

Full Mode

Multi-worker with Redis and Celery for background task processing. For production deployments and teams.

fagoon up --full

---

Using Docker Compose Directly

If you prefer not to use the CLI:

Lite Mode

# docker-compose.yml
services:
  app:
    image: ghcr.io/fagoon-ai/upgrade:2.0.0
    ports:
      - "8000:8000"
      - "3000:3000"
    environment:
      LITE_MODE: "true"
      DATABASE_URL: postgresql+asyncpg://fagoon:fagoon@db:5432/fagoon
      JWT_SECRET: "your-secret-here"
      GEMINI_API_KEY: "your-key-here"
    depends_on:
      db:
        condition: service_healthy

  db:
    image: pgvector/pgvector:pg16
    environment:
      POSTGRES_USER: fagoon
      POSTGRES_PASSWORD: fagoon
      POSTGRES_DB: fagoon
    volumes:
      - fagoon_db:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U fagoon"]
      interval: 5s
      timeout: 5s
      retries: 10

volumes:
  fagoon_db:

docker compose up -d

Full Mode

# docker-compose.full.yml
services:
  app:
    image: ghcr.io/fagoon-ai/upgrade:2.0.0
    ports:
      - "8000:8000"
      - "3000:3000"
    environment:
      LITE_MODE: "false"
      DATABASE_URL: postgresql+asyncpg://fagoon:fagoon@db:5432/fagoon
      REDIS_URL: redis://redis:6379/0
      CELERY_BROKER_URL: redis://redis:6379/1
    depends_on:
      db:
        condition: service_healthy
      redis:
        condition: service_started

  worker:
    image: ghcr.io/fagoon-ai/upgrade:2.0.0
    command: uv run celery -A src.core.celery_app worker -P eventlet -c 50
    environment:
      LITE_MODE: "false"
      DATABASE_URL: postgresql+asyncpg://fagoon:fagoon@db:5432/fagoon
      REDIS_URL: redis://redis:6379/0
      CELERY_BROKER_URL: redis://redis:6379/1
    depends_on:
      app:
        condition: service_healthy

  redis:
    image: redis:7-alpine

  db:
    image: pgvector/pgvector:pg16
    environment:
      POSTGRES_USER: fagoon
      POSTGRES_PASSWORD: fagoon
      POSTGRES_DB: fagoon
    volumes:
      - fagoon_db:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U fagoon"]
      interval: 5s
      timeout: 5s
      retries: 10

volumes:
  fagoon_db:

docker compose -f docker-compose.full.yml up -d

---

Workflow API — Turn Any Workflow Into an API

1. Build a Workflow

Open http://localhost:3000, create a workflow in the visual canvas.

2. Publish as API

Click API Integration → Generate API Key & Slug. Copy the API key (shown only once).

3. Call It From Anywhere

cURL:

curl -X POST http://localhost:8000/api/v1/workflow-api/your-slug/execute \
  -H "X-API-Key: wfapi_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{"input": "Hello from my app"}'

JavaScript:

const response = await fetch('http://localhost:8000/api/v1/workflow-api/your-slug/execute', {
  method: 'POST',
  headers: {
    'X-API-Key': 'wfapi_your_key_here',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({ input: 'Hello from my app' })
});
const data = await response.json();
console.log(data.output);

Python:

import requests

response = requests.post(
    'http://localhost:8000/api/v1/workflow-api/your-slug/execute',
    headers={'X-API-Key': 'wfapi_your_key_here'},
    json={'input': 'Hello from my app'}
)
print(response.json()['output'])

API Response Format

{
  "success": true,
  "execution_id": "uuid",
  "status": "COMPLETED",
  "output": {
    "output_text": "AI generated response here",
    "model": "gemini-2.5-flash",
    "usage": { "input_tokens": 100, "output_tokens": 50 }
  },
  "usage": { "duration_ms": 2500 }
}

---

Environment Variables

Variable	Default	Description
LITE_MODE	true	true for single-process, false for Redis+Celery
DATABASE_URL	auto	PostgreSQL connection string
JWT_SECRET	auto-generated	Secret for JWT tokens
ENCRYPTION_KEY	auto-generated	Fernet key for encrypting secrets
GEMINI_API_KEY	—	Google Gemini API key
OPENAI_API_KEY	—	OpenAI API key
GROQ_API_KEY	—	Groq API key
ANTHROPIC_API_KEY	—	Anthropic API key
REDIS_URL	—	Redis URL (full mode only)
CELERY_BROKER_URL	—	Celery broker URL (full mode only)
WEB_CONCURRENCY	1	Number of uvicorn workers

---

Troubleshooting

"Cannot connect to Docker daemon"

Start Docker Desktop and wait for the whale icon to appear in your menu bar.

"Port 3000/8000 already in use"

Stop any other services on those ports, or change ports in the compose file.

Platform won't start

fagoon logs -f     # Check logs for errors
fagoon down        # Stop everything
fagoon up          # Restart

Reset everything

fagoon down
docker volume rm fagoon_db fagoon_data   # Removes all data
fagoon up

---

Tech Stack

• Backend: Python, FastAPI, SQLAlchemy, PostgreSQL, pgvector
• Frontend: Next.js, React, TailwindCSS, React Flow
• Workflow Engine: Custom DAG executor with 20+ node types
• Database: PostgreSQL with pgvector for embeddings
• Task Queue: Celery + Redis (full mode)
• Container: Docker, Docker Compose

---

License

MIT

---

Links

• Website: fagoonai.com
• GitHub: github.com/Fagoon-AI/upgrade
• PyPI: pypi.org/project/fagoon-upgrade