GitHub - rush86999/atom: Atom Agent, automate your workflows by talking to an AI — and let it remember, search, and handle tasks like a real assistant

8 min read Original article ↗

ATOM Platform

Open-Source AI Agent Platform for Self-Hosted Automation

Developer Note: For technical setup and architecture, see docs/development/overview.md.

Atom Platform

Automate your workflows by talking to an AI — and let it remember, search, and handle tasks like a real assistant.

License Status

What is Atom?

Atom is an open-source, self-hosted AI agent platform that combines visual workflow builders with intelligent LLM-based agents.

Just speak or type your request, and Atom's specialty agents will plan, verify, and execute complex workflows across your entire tech stack.

Key Difference: Atom runs entirely on your infrastructure. Your data never leaves your environment.

Comparing alternatives? See Atom vs OpenClaw for a detailed feature comparison.

Atom vs OpenClaw: Quick Comparison

Aspect Atom OpenClaw
Best For Business automation, multi-agent workflows Personal productivity, messaging workflows
Agent Model Multi-agent system with specialty agents Single-agent runtime
Governance ✅ 4-tier maturity (Student → Autonomous) ❌ No maturity levels
Memory ✅ Episodic memory with graduation validation ✅ Persistent Markdown files
Integrations 46+ business (CRM, support, dev tools) 50+ personal (smart home, media, messaging)
Architecture Python + FastAPI + PostgreSQL/SQLite Node.js + local filesystem
Setup Docker Compose (~15-30 min) Single script (~10-30 min)

Full Comparison →

Architecture

Single-Tenant Deployment

Atom is designed for self-hosted deployment:

  • Simpler Setup: No tenant isolation, no subdomain routing
  • Better Performance: Direct database access without overhead
  • Self-Hosted: Your data never leaves your infrastructure
  • Unlimited Usage: No subscription fees or quota limits

Key Features:

  • Uses user_id for user identification
  • No billing system or quota enforcement
  • Fleet recruitment limited by system resources only
  • All governance, routing, and graduation features work identically

Full Architecture Guide →

Meta-Agent Routing ✨ NEW

Intelligent request routing with governance checks:

  • Intent Classification: Automatically categorize requests as CHAT/WORKFLOW/TASK
  • Governance Checks: Maturity-based permissions for WORKFLOW/TASK routing
  • Auto-Takeover: Propose CHAT alternatives when governance denies actions
  • Dynamic Fleet Recruitment: TASK routing uses FleetAdmiral for specialist recruitment

Quick Start

# Install Atom
pip install atom-os

# Initialize
atom init

# Add your API keys to .env
# OPENAI_API_KEY=sk-...
# MINIMAX_API_KEY=...  (optional, for MiniMax M2.7 support)

# (Optional) Connect to marketplace for commercial specialty skills
# MARKETPLACE_API_TOKEN=at_saas_your_token  # Get from https://atomagentos.com


# Start Atom
atom start

# Open dashboard
open http://localhost:8000

That's it! 🚀

Optional: Connect to Marketplace →

Choose your edition:

  • Personal Edition - Free, single-user, SQLite (default)
  • Enterprise Edition - Multi-user, PostgreSQL, monitoring → pip install atom-os[enterprise]

Full Installation Guide →

Key Features

🎙️ Voice Interface

  • Build complex workflows using just your voice
  • Natural language understanding — no proprietary syntax
  • Real-time feedback as Atom visualizes its reasoning

🤖 Specialty Agents

  • Sales Agent: CRM pipelines, lead scoring, outreach
  • Marketing Agent: Campaigns, social posting, analytics
  • Engineering Agent: PR notifications, deployments, incidents
  • Hive Orchestration: Autonomous multi-agent coordination with Queen (Architectural Design) and King (Execution Orchestration). ✨ NEW
  • Intelligent request routing: Automatic intent classification to route requests to the most capable specialty agent. ✨ NEW
  • Autonomous Workflow Realization: Convert natural language requests into persistent, executable blueprints via the Queen Agent. ✨ NEW
  • Meta-Agent Routing: Governance-gated routing with CHAT/WORKFLOW/TASK intent classification and auto-takeover proposal mode. ✨ NEW (v13.0)
  • Dynamic Fleet Recruitment: FleetAdmiral automatically recruits specialist agents based on task requirements. ✨ NEW (v13.0)
  • Governance: Progress from "Student" to "Autonomous" as they gain trust

🎨 Canvas Presentations & Real-Time Guidance ✨

  • Rich interactive presentations (charts, forms, markdown)
  • Live operation visibility: See exactly what agents are doing
  • Multi-view orchestration (browser, terminal, canvas)
  • Smart error resolution with actionable suggestions
  • Full Details →

🧠 Autonomous Self-Evolution & Graduation ✨ NEW

  • Experience-based learning: Agents store and retrieve past experiences using the Reflection Pool.
  • Recursive Self-Evolution: Autonomous critique-based optimization loop (optimized for MiniMax M2.7).
  • Dual-Trigger Graduation: Skills progress from SUPERVISED to AUTONOMOUS via post-task event hooks and background audits.
  • Dynamic Streak Rule: Promotion based on "Clean Run" streaks (Success + No Intervention + High Compliance).
  • Hybrid storage: PostgreSQL for state + LanceDB for mistake memory.
  • Agent Graduation Guide →

🛡️ Agent Governance System

  • 4-tier maturity-based routing and approval system
  • AI-powered training duration estimation
  • Every action logged, timestamped, and traceable

🔌 Deep Integrations

  • 46+ business integrations: Slack, Gmail, HubSpot, Salesforce, Zendesk
  • 9 messaging platforms: Real-time communication
  • Marketplace Connection: Access 5,000+ community skills and agent templates ✨ NEW
  • Use /run, /workflow, /agents from your favorite chat app

🔍 Knowledge Graph & GraphRAG ✨ NEW

  • Recursive Knowledge Retrieval: Higher-order reasoning via BFS graph traversal.
  • Canonical Anchoring: Link graph nodes to concrete database records (Users, Tickets, Formulas).
  • Bidirectional Sync: Update entity metadata directly from the Graph UI.
  • Visual Explorer: Interactive D3-powered graph visualization at /graph.
  • GraphRAG Documentation →

🌐 Community Skills & Package Marketplace ✨

  • 5,000+ OpenClaw/ClawHub skills: Import community-built tools
  • Skill Marketplace: PostgreSQL-based with search, ratings, categories
  • Enterprise security: LLM-powered scanning (21+ malicious patterns) with hazard sandbox
  • Skills Registry: Import with status tracking (Untrusted → Active → Banned)
  • Dynamic Skill Loading: importlib-based hot-reload with watchdog monitoring
  • Skill Composition: DAG workflows with parallel execution and rollback
  • Auto-Installation: Python + npm with conflict detection and automatic rollback
  • Python Package Support: NumPy, Pandas, 350K+ PyPI packages with vulnerability scanning
  • npm Package Support: Lodash, Express, 2M+ npm packages with governance
  • Supply Chain Security: Typosquatting detection, dependency confusion prevention, postinstall malware blocking
  • Community Skills Guide →

🔍 Browser & Device Automation

  • Browser automation via CDP (scraping, form filling)
  • Device control (camera, location, notifications)
  • Maturity-governed for security

Installation Options

🐳 Docker (5 minutes)

git clone https://github.com/rush86999/atom.git
cd atom
cp .env.personal .env
docker-compose -f docker-compose-personal.yml up -d

🚀 DigitalOcean (1-Click Deploy)

Launch Atom on DigitalOcean App Platform with one click:

Deploy to DO

See Cloud Deployment Guide →

💻 Native (10 minutes)

git clone https://github.com/rush86999/atom.git
cd atom
cd backend && python3 -m venv venv && source venv/bin/activate
pip install -r requirements.txt
cd ../frontend-nextjs && npm install
cp .env.personal .env
# Start backend: cd backend && python -m uvicorn main_api_app:app --reload
# Start frontend: cd frontend-nextjs && npm run dev

Marketplace (Commercial Service)

The Atom Agent OS includes a commercial marketplace for discovering and installing:

  • Agents - Pre-configured AI agents for specific tasks
  • Domains - Specialist knowledge domains (Sales, Engineering, Support)
  • Components - Reusable canvas workflow components
  • Skills - Agent capabilities and integrations

Connection to atomagentos.com

The Marketplace requires an active connection to atomagentos.com and a valid API token. To enable marketplace features:

  1. Get Your API Token

    • Visit atomagentos.com
    • Sign up and navigate to Settings → API Tokens
    • Generate and copy your API token

  2. Configure Environment

    # Add to .env
ATOM_SAAS_API_URL=https://atomagentos.com/api/v1
ATOM_SAAS_API_TOKEN=your_api_token_here
MARKETPLACE_ENABLED=true

  3. Restart Atom

Commercial Terms

Important: Marketplace items are proprietary content licensed under the atomagentos.com Terms of Service.

  • Service Dependency: The Marketplace feature requires an active connection to atomagentos.com
  • Proprietary Content: All marketplace items (skills, agents, domains, components) are proprietary
  • Redistribution Prohibited: Marketplace items may not be redistributed, resold, or shared outside the platform
  • API Token Required: Access to the Marketplace Service requires a valid API token from atomagentos.com
  • Service Availability: The Marketplace Service is provided "as is" and availability is not guaranteed

License Distinction:

  • Core Platform: AGPL v3 (open source) - Agent runtime, cognitive architecture, governance
  • Marketplace Service: Commercial/Proprietary - Requires atomagentos.com connection and API token

For full legal terms, see LICENSE.md - Marketplace Commercial Appendix.

Documentation

Example Use Cases

Department Scenario
Sales New lead in HubSpot → Research → Score → Notify Slack
Finance PDF invoice in Gmail → Extract → Match QuickBooks → Flag discrepancies
Support Zendesk ticket → Analyze sentiment → Route urgent → Draft response
HR New employee in BambooHR → Provision → Invite → Schedule orientation

Security & Privacy

  • Self-Hosted Only: Your data never leaves your environment
  • BYOK: Bring your own OpenAI, Anthropic, Gemini, DeepSeek, or MiniMax keys
  • Encrypted Storage: Sensitive data encrypted at-rest (Fernet)
  • Audit Logs: Every agent action logged and traceable
  • Human-in-the-Loop: Configurable approval policies
  • Package Security: Vulnerability scanning (pip-audit, Safety), postinstall script blocking, container isolation
  • Supply Chain Protection: Typosquatting detection, dependency confusion prevention, malicious pattern detection
  • Comprehensive Testing: 495+ tests with 99%+ pass rate, E2E validation with real services ✨
  • AI-Enhanced Bug Discovery: Automated fuzzing, property-based testing, chaos engineering ✨ NEW
  • Stress Testing: Load testing with k6 (10/50/100 concurrent users), network simulation, failure injection ✨ NEW

Documentation

User Guides ⭐

Core Features

Testing & Quality ✨ NEW

Platform

Advanced

Complete Documentation Index → | Reorganization Plan →

Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

Support