Spark

**Because magic shouldn't be complicated. **

Spark is an automagion engine that seamlessly integrates with multiple LangFlow instances. Deploy AI-driven flows, schedule one-time or recurring tasks, and monitor everything with minimal fuss—no coding required.

🔗 Ecosystem

• AutoMagik Agents: Develop production-level AI agents
• AutoMagik UI: Create agents using natural language with our dedicated UI

🚀 Installation

Spark provides two setup options:

Prerequisites

• Linux-based system (Ubuntu/Debian recommended)
• Docker and Docker Compose (automatically installed on Ubuntu/Debian if not present)

Local Production Setup

For a production-ready local environment:

./scripts/setup_local.sh

Development Setup

For development with PostgreSQL and Redis Docker containers:

./scripts/setup_dev.sh

What Happens During Setup

Both setup scripts will:

• Create necessary environment files
• Install Docker if needed (on Ubuntu/Debian)
• Set up all required services
• Install the CLI tool (optional)
• Guide you through the entire process

After Installation

You'll have access to:

• Spark API: Running at http://localhost:8883
• PostgreSQL Database: Available at localhost:15432
• Worker Service: Running and ready to process tasks
• CLI Tool: Installed (if chosen during setup)

Verifying Your Installation

The setup automatically verifies all services, but you can also check manually:

# Access API documentation
open http://localhost:8883/api/v1/docs  # Interactive Swagger UI
open http://localhost:8883/api/v1/redoc # ReDoc documentation

# List flows (requires CLI installation)
source .venv/bin/activate
automagik-spark flow list

🧩 System Components

• API Server: Handles all HTTP requests and core logic
• Worker: Processes tasks and schedules
• Database: PostgreSQL with all required tables automatically created
• LangFlow (optional): Visual flow editor for creating AI workflows
• CLI Tool (optional): Command-line interface for managing flows and tasks

🏗️ System Architecture

flowchart LR
    subgraph Services
      DB[PostgreSQL]
      LF1[LangFlow Instance 1]
      LF2[LangFlow Instance 2]
    end
    subgraph Spark
      CLI[CLI]
      API[API Server]
      CW[Celery Worker]
      W[Worker]
    end
    API -- uses --> DB
    API -- triggers --> CW
    W -- processes --> API
    API -- integrates with --> LF1
    API -- integrates with --> LF2
    CLI -- controls --> API
    API -- has UI --> UI[Automagik UI]

Loading

Core Components Explained

• API: Core service handling requests and business logic
• Worker: Processes tasks and schedules
• CLI: Command-line tool for managing flows and tasks
• PostgreSQL: Stores flows, tasks, schedules, and other data
• LangFlow: Optional service for creating and editing flows

📚 API Documentation

For complete API documentation, visit:

• Swagger UI: http://localhost:8883/api/v1/docs
• ReDoc: http://localhost:8883/api/v1/redoc

🛠️ Next Steps

1. If you installed LangFlow, visit http://localhost:17860 to create your first flow
2. Use the API at http://localhost:8883/api/v1/docs to manage your flows and tasks
3. Try out the CLI commands with automagik-spark --help
4. Monitor task execution through logs and API endpoints

📊 Telemetry

Spark collects anonymous usage analytics to help improve the project. This data helps us understand which features are most useful and prioritize development efforts.

What We Collect

• Command usage and performance metrics
• API endpoint usage patterns
• Workflow execution statistics
• System information (OS, Python version)
• Error rates and types

What We DON'T Collect

• Personal information or credentials
• Actual workflow data or content
• File paths or environment variables
• Database connection strings or API keys

How to Disable Telemetry

Environment Variable:

export AUTOMAGIK_SPARK_DISABLE_TELEMETRY=true

CLI Commands:

# Disable permanently
automagik-spark telemetry disable

# Check status
automagik-spark telemetry status

# See what data is collected
automagik-spark telemetry info

# Use --no-telemetry flag for single session
automagik-spark --no-telemetry <command>

Opt-out File:

touch ~/.automagik-no-telemetry

Telemetry is automatically disabled in CI/testing environments.

🗺️ Roadmap

Spark's future development focuses on:

• TBA

---

Spark: Bringing AI Automation to Life