Automagik Multi-Agent Framework - Extensible Agent Creation System
Project description
๐ง Automagik Hive
YAML-powered AI agent orchestrator: Build teams, not code.
The framework that turns agent coordination from nightmare to magic
โก 5-Minute Magic โข ๐ฏ Why Hive? โข ๐ Examples โข ๐๏ธ Architecture
โก 5-Minute Magic
Watch your wish come true: From zero to AI project management team in 5 minutes
Step 1: Prerequisites (30 seconds)
# One command installs everything you need (UV, Python 3.12+, Docker, Make)
curl -fsSL https://raw.githubusercontent.com/namastexlabs/automagik-hive/main/scripts/install-predeps.sh | bash
Step 2: Wake Up Genie (30 seconds)
# Start Genie in your workspace folder
uvx automagik-hive ./my-workspace
# Genie is now running on http://localhost:38886 and ready for your wish!
Step 3: Magic Happens (4 minutes)
You tell Genie your wish via chat or web interface:
You: "I need a team to manage my software project - planning, coding, testing, and deployment"
Genie: "Perfect! I understand you need a project management team.
Let me create that for you right now...
โจ Creating ./project-management-team/ with:
๐ฏ Project Planner - Requirements analysis and roadmaps
๐ป Senior Developer - Full-stack development and architecture
๐งช QA Specialist - Testing strategy and quality assurance
๐ DevOps Engineer - Deployment and infrastructure
๐ Launching your team now on port 8887...
Your project management team is live! What's your first project?"
You: "Build a todo application with React frontend and Python backend"
Genie: "Excellent! I'm coordinating your new team:
- Planner: Creating technical specification for React/Python todo app...
- Developer: Setting up project structure with best practices...
- QA: Preparing comprehensive testing strategy...
- DevOps: Configuring deployment pipeline...
โจ Your todo app project is now actively being built by your AI team!
Monitor progress at: http://localhost:8887"
What just happened? In 5 minutes you got:
- ๐งโโ๏ธ Genie Consultation - Personal AI assistant understanding your needs
- ๐ค Custom Agent Team - Genie created and launched the perfect team for you
- ๐ Active Project - Real todo app being built by your AI team
- ๐ Dual-Instance Architecture - Genie (port 38886) + Your Team (port 8887)
- ๐ Production Pipeline - Testing, deployment, monitoring all configured
The Magic: You made a wish to Genie, Genie created your perfect team, and your project is running!
๐ฏ Why Automagik Hive?
The YAML Revolution: Stop writing orchestration code. Start writing configurations.
| Feature | Automagik Hive | LangChain | AutoGPT | CrewAI |
|---|---|---|---|---|
| Agent Definition | YAML configs | Python code | Prompts | Python classes |
| Team Orchestration | Built-in routing | Custom chains | Single agent | Manual roles |
| External Integration | Native MCP | Plugin system | Limited | Basic |
| Development Speed | Minutes | Hours | Days | Hours |
| Genie Assistant | โ Built-in | โ | โ | โ |
| External Agents | โ Dynamic loading | โ | โ | โ |
The Difference: While others make you a programmer, we make you a conductor. ๐ผ
๐ Real-World Magic
๐ผ Code Review Automation
# code-review-team.yaml
agents:
- name: quality-inspector
role: "Analyze code for bugs, security issues, and best practices"
- name: test-generator
role: "Create comprehensive test suites for new code"
- name: documentation-writer
role: "Generate clear documentation and comments"
workflow:
- quality-inspector โ test-generator โ documentation-writer
One command: uvx automagik-hive ./code-review-team.yaml
Result: Automated code reviews that catch bugs, generate tests, and write docs.
๐ฑ Content Pipeline
# viral-content-team.yaml
agents:
- name: trend-researcher
role: "Monitor Reddit, Twitter, HN for viral topics"
- name: content-creator
role: "Generate engaging content from trending topics"
- name: distributor
role: "Post content across social platforms"
triggers:
- schedule: "0 */6 * * *" # Every 6 hours
- webhook: "/new-trend"
Result: Autonomous content creation pipeline that finds trends and creates viral content.
๐ข Enterprise Workflow
# support-automation.yaml
agents:
- name: ticket-classifier
role: "Categorize and prioritize customer tickets"
- name: solution-finder
role: "Search knowledge base and suggest solutions"
- name: escalation-manager
role: "Route complex issues to human experts"
integrations:
- slack_notifications
- jira_updates
- customer_database
Result: 80% of support tickets handled automatically with intelligent escalation.
๐๏ธ How It Works
The Magic Behind The Scenes
Your YAML โ Automagik Hive โ Production Magic
[Agent Configs] โ [Dynamic Loading] โ [Intelligent Routing] โ [Task Execution]
โ โ โ โ
Simple YAML Runtime Discovery Smart Coordination Real Results
Core Philosophy: Configuration over code, orchestration over implementation.
๐งโโ๏ธ Meet Genie - Your AI Development Partner
Genie isn't just an agent - it's your development companion that:
- Helps you design better agent teams through conversation
- Writes YAML configs based on your natural language requirements
- Debugs issues when agents aren't coordinating properly
- Suggests optimizations for better performance
- Learns from your patterns to make future suggestions smarter
Example conversation with Genie:
You: "I need agents to monitor our app for issues and fix them automatically"
Genie: "I'll create a monitoring team with alerting, diagnosis, and auto-fix agents.
Let me generate the YAML config and set up the monitoring triggers..."
๐ง Installation & Setup
Option 1: Instant Start (Recommended)
# Wake up Genie in your workspace
uvx automagik-hive ./my-workspace
# Tell Genie what kind of team you need
# Genie will create and launch the perfect team for you
# Visit: http://localhost:38886 to chat with Genie
Option 2: Development Mode
# Clone for framework development
git clone https://github.com/namastexlabs/automagik-hive.git
cd automagik-hive
# Install with UV (recommended)
uv sync
uv run python -m automagik_hive.cli ./examples/starter-team
Option 3: MCP Integration
Connect to your existing Claude Desktop workflow:
// ~/.claude_desktop_config/config.json
{
"mcpServers": {
"automagik-hive": {
"command": "uvx",
"args": ["automagik-hive", "--mcp-server"],
"env": {
"HIVE_PROJECT_PATH": "/path/to/your/agents"
}
}
}
}
๐ Agent Development
Creating Your First Agent
# agents/my-specialist.yaml
name: data-analyst
version: 1.0.0
capabilities:
- data_analysis
- visualization
- reporting
prompt: |
You are a world-class data analyst specializing in business intelligence.
When given data, you:
1. Analyze patterns and trends
2. Create clear visualizations
3. Provide actionable insights
4. Format results for business stakeholders
tools:
- pandas_query
- chart_generator
- report_builder
memory:
type: persistent
scope: user_session
That's it! Your agent is automatically discovered and integrated into the team.
Team Coordination
# teams/analytics-team.yaml
name: business-intelligence
version: 1.0.0
members:
- data-analyst
- visualization-expert
- report-writer
routing:
default: data-analyst
complex_analysis: [data-analyst, visualization-expert]
final_reports: report-writer
workflows:
monthly_report:
trigger: "0 0 1 * *" # First day of each month
steps:
- data-analyst: "Analyze last month's metrics"
- visualization-expert: "Create executive dashboards"
- report-writer: "Generate monthly business report"
๐ Advanced Features
๐ Dynamic Agent Loading
- Hot Reload: Add new agents without restarting
- Version Management: A/B test different agent versions
- Dependency Resolution: Agents automatically discover their tools
๐ง Intelligent Memory
- Persistent Context: Agents remember across sessions
- Shared Knowledge: Teams build collective intelligence
- Learning Loops: Performance improves over time
๐ Universal Integration
- MCP Protocol: Connect to any MCP-compatible tool
- REST APIs: Standard HTTP interfaces for everything
- Webhooks: Real-time event handling
- Databases: Native PostgreSQL with vector search
๐ก๏ธ Production Ready
- Authentication: Multi-layer security with API keys
- Monitoring: Full observability with metrics and logs
- Scaling: Horizontal scaling with Docker/Kubernetes
- Error Handling: Graceful degradation and recovery
๐จ Example Gallery
Starter Templates
- Code Assistant - Development workflow automation
- Content Creator - Social media management
- Data Pipeline - ETL and analysis automation
- Customer Support - Ticket handling and escalation
- DevOps Helper - Infrastructure monitoring and deployment
Community Examples
- E-commerce Bot - Product recommendations and customer service
- Research Assistant - Academic paper analysis and summarization
- Trading Bot - Market analysis and automated trading
- Content Moderator - Social platform content filtering
๐ค Community & Support
๐ Join the Hive
- Discord Community - Real-time help and collaboration
- GitHub Discussions - Feature requests and architecture discussions
- Agent Marketplace - Share and discover community agents
๐ Documentation
- Getting Started Guide - Complete onboarding
- Agent Development - Build custom agents
- Team Orchestration - Coordinate agent interactions
- Production Deployment - Scale to enterprise
๐ Learning Resources
- Video Tutorials - Step-by-step walkthroughs
- Blog Posts - Deep dives and case studies
- Webinars - Live demos and Q&A sessions
๐ What's Next?
For New Users
- Try the 5-minute demo - See the magic in action
- Browse examples - Find templates for your use case
- Join Discord - Get help from the community
For Developers
- Read the architecture docs - Understand the magic
- Create your first agent - Build something custom
- Contribute to the project - Help make the magic better
For Enterprises
- Schedule a demo - See enterprise features
- Review security docs - Understand compliance
- Contact sales - Discuss your specific needs
๐ค Contributing
Help make agent orchestration magical for everyone!
# Get started with development
git clone https://github.com/namastexlabs/automagik-hive.git
cd automagik-hive
uv sync
uv run python -m automagik_hive.cli --dev
# All contributions include Genie collaboration!
git commit -m "feat: new agent capability
Co-Authored-By: Automagik Genie <genie@namastex.ai>"
Ways to contribute:
- ๐ Bug Reports - Help us squash issues
- ๐ก Feature Ideas - Share your vision for better orchestration
- ๐ค New Agents - Build reusable agents for the community
- ๐ Documentation - Make the magic more accessible
- ๐จ Examples - Show off creative use cases
๐ License
MIT License - build incredible things and share the magic! See LICENSE for details.
๐งโโ๏ธ "Your wish is my command" - Genie
Ready to orchestrate some magic?
๐ Start Your First Agent Team
Built with โจ by the Automagik Team & Genie
Where YAML meets magic, and agents become teams.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file automagik_hive-0.1.0a36.tar.gz.
File metadata
- Download URL: automagik_hive-0.1.0a36.tar.gz
- Upload date:
- Size: 671.8 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.12.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
791be0f8c71a5260284396311ab9393a29db83f4c697a2a4d9af85c5d7a8b02d
|
|
| MD5 |
bedb435e90e53597dd921576a0fce2c4
|
|
| BLAKE2b-256 |
6e1aed6b4f200bbaaade2ca1cf80532c89a836daf812c0e843b13363cad9f57d
|
File details
Details for the file automagik_hive-0.1.0a36-py3-none-any.whl.
File metadata
- Download URL: automagik_hive-0.1.0a36-py3-none-any.whl
- Upload date:
- Size: 434.3 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.12.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
730703eb98edeeb6170317ec28cc09e1a1a7f1781bb18eb09eba068b19556212
|
|
| MD5 |
0343183cba6ccde45b69ba392105c2f9
|
|
| BLAKE2b-256 |
7b551f9d16f9d2315dc85d4d4aaee67e66be72404088d84d8fd1e358749bbc94
|