Skip to content

Your AI Engineering Team in a Box. AutoMates is a role-based multi-agent framework that transforms Large Language Models into a structured engineering department.

License

Notifications You must be signed in to change notification settings

Nate-Vish/Auto-Mates

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AutoMates.AI 👾

Your AI Engineering Team in a Box.

"We believe everyone's imagination deserves to become their creation."


🚀 Quick Start

1. Describe Your Project

Edit Dashboard/Project_Description.md with your vision.

2. Set Your Rules

Edit Dashboard/Rules.md with any constraints (tech stack, style, etc.).

3. Activate an Agent

  1. Open a terminal in the AutoMates folder
  2. Activate an AI model (Claude Code, Gemini CLI, your local model, etc.)
  3. Type you are [Agent_Name]: + drag the wanted identity file to terminal + Enter
  4. Agent pops up after reading his identity, his memory, and the dashboard. Ready to rumble.
you are BrainStorm: [drag Agents/BrainStorm/BrainStorm_Identity.md here]

4. Run Multiple Agents

Open a few terminals, activate agents in each one. They collaborate through files in Dashboard/Work_Space/.


💡 Tip: Use Fetcher to collect educational material for the agents. Let them read it before they start working on a task (they'll find it in Library, don't worry).

💡 Hint: Every agent can create a file for other agents to read and work by.

📖 Example: Fetcher can read a Blueprint that Planner made, go fetch some sources and create Builder_Study.md, then Builder learns like a pro and starts writing some fine code.

🧠 Try this: Got a quick idea while working? Just tell BrainStorm 💡 IDEA: [your idea] — he'll log it instantly and keep working. No flow disruption, ideas never lost.


⚙️ How It Works

👥 The Team (8 Agents)

Agent Role What They Do
🧠 BrainStorm Ideator Explores ideas, solves creative blocks
📐 Planner Architect Creates technical blueprints
🔨 Builder Engineer Writes code following blueprints
🕵️ Checker Auditor Reviews for bugs, security, quality
⚖️ Legal Compliance Checks licenses, privacy, governance
📦 GitDude Release Manager Version control, security scanning
📚 Fetcher Librarian Gathers knowledge, organizes sources
🎼 Orca Orchestrator Modifies agents, creates new ones, manages team structure

📂 The Three Zones

AutoMates/
├── Agents/                    # Where agents live
│   ├── BrainStorm/
│   ├── Planner/
│   ├── Builder/
│   ├── Checker/
│   ├── Legal/
│   ├── GitDude/
│   └── Orca/
│
├── Library/                   # Where information is stored
│   ├── Fetcher/              # He lives here near his Sources
│   └── Sources/              # Organized knowledge base
│
└── Dashboard/                 # Where agents work together
    ├── Project_Description.md
    ├── Rules.md
    ├── Work_Space/           # The fun happens here
    └── Version_Control/

Agents/ — Where the agents live, including their identity and memory.

Library/ — Where information is stored and organized by Fetcher (he lives there near his Sources).

Dashboard/ — The place where all agents work together on your tasks and manage the versions for you (the fun happens there).

🧠 Agent Memory

Each agent remembers past sessions at Agents/[Name]/Memory_Logs/:

  • Lessons.md — Patterns that worked, mistakes to avoid
  • Preferences.md — How you like things done
  • Sessions/ — Conversation history
  • Checkpoint.md — Save/resume complex tasks

🔮 Roadmap

Current (v1.x):

  • 8 specialized agents with persistent memory
  • LEARN FIRST protocol with Study Files
  • File-based coordination via Work_Space
  • Orca for team customization and new agent creation

Coming (v2.0+):

  • Orchestrator Automation (Orca managing workflows automatically)
  • Agent Factory (generate specialized teams)
  • CLI Package (automates command)

⚖️ Legal, License & Links

🛡️ You Are the Pilot-in-Command

  • Your Responsibility: You control what agents do. You are responsible for URLs fetched and code generated.
  • Third-Party Tools: AutoMates uses r.jina.ai as a free utility. Users engaging in high-volume or commercial use are responsible for obtaining their own API key to comply with Jina AI's terms.
  • AI Disclaimer: AI-generated content can contain errors. Always review and test all output.
  • Copyright: Respect intellectual property when using Fetcher.

🔒 Privacy

  • Third-Party APIs: Prompts are sent to AI providers (Anthropic, OpenAI, Google). See their privacy policies.
  • Local Storage: AutoMates works locally. We do not store your code or data.
  • No Telemetry: Everything stays on your machine.

🐋 Docker Recommended

We recommend running each project in its own Docker container for isolation and safety.

📄 License

MIT License © 2026 AutoMates.AI — See LICENSE

🔗 Links

About

Your AI Engineering Team in a Box. AutoMates is a role-based multi-agent framework that transforms Large Language Models into a structured engineering department.

Resources

License

Contributing

Stars

Watchers

Forks

Packages

No packages published

Contributors 2

  •  
  •