The Automation Engine That Never Sleeps

🎯 Turn Reactive AI into Proactive Intelligence
Schedule workflows, orchestrate agents, automate everything—
so your AI works while you sleep

Features • Quick Start • Roadmap • Development • Contributing

---

🚀 What is Automagik Spark?

Automagik Spark is the automation engine that gives your AI agents a sense of time. While LangFlow and Hive create powerful workflows and agents, Spark makes them proactive—running tasks on schedules, triggering actions automatically, and working 24/7 without manual intervention.

Think of Spark as the alarm clock, calendar, and scheduler for your entire AI infrastructure.

💡 The Problem We Solve

The Challenge: Your AI workflows are brilliant but passive. They sit idle until you remember to trigger them:

• Daily reports that you forget to generate
• Monitoring tasks that need constant attention
• Routine automations that require manual kicks
• Agents that only respond when asked

Our Solution: Spark transforms passive workflows into active systems that work autonomously on your schedule.

✅ How Spark Changes the Game

Before Spark:

• ❌ Manually trigger workflows every time
• ❌ Set reminders to run routine tasks
• ❌ AI that only reacts when prompted
• ❌ Miss opportunities because agents weren't watching

After Spark:

• ✅ Workflows run automatically on schedule
• ✅ Set it once, runs forever
• ✅ AI that takes initiative proactively
• ✅ Never miss a beat with 24/7 automation

---

🌟 Key Features

• ⏰ Smart Scheduling: Cron expressions, intervals, or one-time runs—schedule anything
• 🔄 Multi-Source Integration: Connect unlimited LangFlow and Hive instances
• 🤖 Workflow Orchestration: Sync and manage flows/agents from a unified control center
• 👷 Distributed Workers: Scale task execution with Celery-powered workers
• 📊 Real-time Monitoring: Track every task execution with detailed logs and status
• 🎯 MCP Native: Control Spark programmatically via Model Context Protocol
• 🔒 Privacy-First Telemetry: Anonymous usage analytics with full opt-out control
• 📡 REST API: Full-featured API with interactive Swagger documentation
• 🛠️ Powerful CLI: Manage everything from the command line
• 🐳 Flexible Deployment: Docker, PM2, or bare metal—your choice

---

🎭 Real-World Use Cases

For Solo Developers

# Morning briefing agent runs daily at 9 AM
automagik-spark schedule create my-briefing-agent "0 9 * * *"

# Database backup runs every Sunday at 2 AM
automagik-spark schedule create backup-flow "0 2 * * 0"

# Monitor API health every 5 minutes
automagik-spark schedule create health-check "*/5 * * * *"

For Teams

• Customer Support: Auto-respond to tickets during off-hours
• DevOps: Automated health checks and incident response
• Data Teams: Scheduled ETL pipelines and report generation
• Marketing: Automated content generation and social posts

For Enterprises

• Compliance: Scheduled audits and compliance checks
• Operations: Proactive monitoring and alerting systems
• Analytics: Automated data processing and insights
• Finance: Scheduled reconciliation and reporting

---

🏗️ How Spark Works

Architecture Overview

graph TB
    subgraph "Your AI Infrastructure"
        LF[LangFlow Instances]
        HV[Automagik Hive]
        OM[Automagik Omni]
    end

    subgraph "Spark Engine"
        API[API Server<br/>FastAPI]
        WK[Workers<br/>Celery]
        DB[(PostgreSQL<br/>State & Tasks)]
        RD[(Redis<br/>Queue)]
    end

    subgraph "Interfaces"
        CLI[CLI Tool]
        MCP[MCP Server]
        REST[REST API]
    end

    LF <-->|Sync Flows| API
    HV <-->|Sync Agents| API
    API -->|Queue Tasks| RD
    RD -->|Execute| WK
    WK -->|Trigger| LF
    WK -->|Trigger| HV
    WK -->|Notify| OM
    API <-->|Store State| DB
    CLI -->|Commands| API
    MCP -->|Control| API
    REST -->|HTTP| API

Loading

The Spark Lifecycle

1. Connect: Link your LangFlow/Hive instances
2. Sync: Discover all available workflows and agents
3. Schedule: Define when and how often tasks should run
4. Execute: Workers trigger workflows automatically
5. Monitor: Track execution history and results
6. Scale: Add more workers as your automation grows

Example: Daily Report Automation

# 1. Add your LangFlow instance
automagik-spark source add http://langflow:7860 YOUR_API_KEY

# 2. Sync available flows
automagik-spark workflow sync "daily-sales-report"

# 3. Schedule it to run every weekday at 8 AM
automagik-spark schedule create WORKFLOW_ID "0 8 * * 1-5"

# That's it! Your report generates automatically every morning

---

📦 Quick Start

Prerequisites

• Python 3.12+ (we use the latest async features)
• PostgreSQL 12+ (for persistent state)
• Redis 6+ (for task queuing)
• Optional: Docker & Docker Compose (for containerized setup)

One-Command Installation

For Production (Docker)

git clone https://github.com/namastexlabs/automagik-spark.git
cd automagik-spark
./scripts/setup_local.sh

For Development

git clone https://github.com/namastexlabs/automagik-spark.git
cd automagik-spark
./scripts/setup_dev.sh

What Gets Installed

After setup completes, you'll have:

• API Server: http://localhost:8883
• Interactive Docs: http://localhost:8883/api/v1/docs
• PostgreSQL: localhost:15432
• Redis: localhost:6379
• CLI Tool: automagik-spark command
• Workers: Running and ready to execute tasks

Verify Installation

# Check API health
curl http://localhost:8883/api/v1/health

# Try CLI commands
source .venv/bin/activate
automagik-spark --help

# View API documentation
open http://localhost:8883/api/v1/docs

---

🎮 Using Spark

Via CLI (Recommended for Developers)

# Add a workflow source
automagik-spark source add https://my-langflow.com API_KEY_HERE

# List all available workflows
automagik-spark workflow list

# Sync a specific workflow
automagik-spark workflow sync "email-processor"

# Create a schedule (runs daily at midnight)
automagik-spark schedule create WORKFLOW_ID "0 0 * * *"

# List all schedules
automagik-spark schedule list

# View task execution history
automagik-spark task list

Via REST API (For Integrations)

# Add a source
curl -X POST http://localhost:8883/api/v1/sources \
  -H "Content-Type: application/json" \
  -d '{"url": "https://langflow.example.com", "api_key": "..."}'

# Create a schedule
curl -X POST http://localhost:8883/api/v1/schedules \
  -H "Content-Type: application/json" \
  -d '{"workflow_id": "...", "cron": "0 0 * * *"}'

Via MCP (From AI Coding Agents)

Spark is available as an MCP tool in Automagik Tools:

# Install MCP server
uvx automagik-tools hub

# Now use natural language from Claude Code, Cursor, etc:
# "Schedule my daily-report workflow to run every morning at 9am"

---

🔧 Configuration

Environment Variables

# Database
DATABASE_URL=postgresql://user:pass@localhost:15432/spark

# Redis
REDIS_URL=redis://localhost:6379

# API
API_HOST=0.0.0.0
API_PORT=8883

# Telemetry (optional)
AUTOMAGIK_SPARK_DISABLE_TELEMETRY=false

Cron Expression Examples

"*/5 * * * *"      # Every 5 minutes
"0 * * * *"        # Every hour
"0 0 * * *"        # Daily at midnight
"0 9 * * 1-5"      # Weekdays at 9 AM
"0 0 * * 0"        # Every Sunday at midnight
"0 0 1 * *"        # First day of every month

---

📊 Monitoring & Telemetry

Built-in Monitoring

# View recent task executions
automagik-spark task list --limit 20

# Check worker status
automagik-spark worker status

# View system health
curl http://localhost:8883/api/v1/health

Privacy-Focused Telemetry

Spark collects anonymous usage metrics to improve the product:

What we collect:

• Command usage frequency
• API endpoint usage patterns
• Workflow execution statistics
• Error rates (no error details)

What we DON'T collect:

• Personal information
• Workflow content or data
• API keys or credentials
• File paths or environment variables

Disable telemetry anytime:

# Environment variable
export AUTOMAGIK_SPARK_DISABLE_TELEMETRY=true

# CLI command
automagik-spark telemetry disable

# Opt-out file
touch ~/.automagik-no-telemetry

---

🌐 Integration with Automagik Suite

Spark is the heartbeat of the Automagik ecosystem:

• Automagik Hive: Schedule multi-agent workflows
• Automagik Omni: Send notifications on schedule
• Automagik Forge: Trigger task execution
• Automagik Tools: Control Spark via MCP
• LangFlow: Schedule visual AI workflows

---

🛠️ Development

Setup Development Environment

# Clone and setup
git clone https://github.com/namastexlabs/automagik-spark.git
cd automagik-spark
./scripts/setup_dev.sh

# Activate virtual environment
source .venv/bin/activate

# Run tests
pytest

# Check code quality
ruff format . && ruff check . && mypy .

# Run API server in dev mode
python -m automagik_spark.api

Project Structure

automagik-spark/
├── automagik_spark/          # Main package
│   ├── api/                  # FastAPI application
│   ├── workers/              # Celery workers
│   ├── models/               # SQLAlchemy models
│   ├── services/             # Business logic
│   └── cli/                  # CLI commands
├── tests/                    # Test suite
├── scripts/                  # Setup and utility scripts
└── docs/                     # Documentation

Running Tests

# Run all tests
pytest

# Run with coverage
pytest --cov=automagik_spark --cov-report=html

# Run specific test file
pytest tests/test_scheduler.py

# Run with verbose output
pytest -v

---

🗺️ Roadmap

Completed ✅

• Multi-source workflow management (LangFlow + Hive)
• Cron-based scheduling with interval support
• Distributed worker architecture with Celery
• REST API with Swagger documentation
• Powerful CLI tool for workflow management
• MCP server integration via Automagik Tools
• Privacy-focused telemetry system

Next Up 🚀

• Step-by-step Workflows: Break complex automations into discrete steps
• Step Outputs: Pass data between workflow steps seamlessly
• Natural Language Scheduling: "Run every morning" instead of cron syntax
• /wish System: AI assistant that helps build workflows
• Improved DX: Simpler setup and instance management
• Workflow Dependencies: Trigger workflows based on other completions
• Omni Integration: Deep integration with messaging platform
• Forge Integration: Trigger tasks from Kanban board
• Conditional Execution: Run workflows based on conditions
• Retry Policies: Auto-retry failed tasks with backoff
• Workflow Marketplace: Share and discover automation templates
• Visual Workflow Builder: Drag-and-drop scheduling interface

Future Vision 🌟

• Advanced workflow orchestration with branching logic
• Multi-tenant support for enterprise deployments
• Real-time collaboration on workflow management
• AI-powered workflow optimization suggestions
• Integration with major observability platforms
• Cloud-hosted version for zero-setup deployments

---

🤝 Contributing

We love contributions! However, to maintain project coherence:

1. Discuss First: Open an issue before starting work
2. Align with Roadmap: Ensure changes fit our vision
3. Follow Standards: Match existing code patterns (async/await, type hints)
4. Test Thoroughly: Include tests for new features (>70% coverage)
5. Document Well: Update docstrings and documentation
6. Quality Checks: Run ruff format . && ruff check . && mypy .
7. Conventional Commits: Use proper commit messages with co-author:

   feat: add step-based workflow execution
   
   Co-authored-by: Automagik Genie 🧞 <[email protected]>
   

See CONTRIBUTING.md for detailed guidelines.

---

🙏 Acknowledgments

Special thanks to:

• The LangFlow team for building an amazing visual workflow platform
• The Celery project for robust distributed task execution
• All our early adopters and contributors who helped shape Spark
• The open-source community for inspiration and support

---

📄 License

MIT License - see LICENSE file for details.

---

🔗 Links

• GitHub: github.com/namastexlabs/automagik-spark
• Discord: discord.gg/xcW8c7fF3R
• Twitter: @automagikdev
• DeepWiki Docs: deepwiki.com/namastexlabs/automagik-spark

---

🚀 Stop manually triggering workflows. Start automating everything.
Spark - The Automation Engine That Never Sleeps ⚡

Star us on GitHub • Join our Discord

Made with ❤️ by Namastex Labs
AI that elevates human potential, not replaces it