Skip to content

Unified multi-agent orchestration layer for Claude Code with parallel execution, session persistence, and natural language routing.

License

Notifications You must be signed in to change notification settings

AgriciDaniel/ace-claude-code

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

A.C.E. - Agent Creates Everything

A unified multi-agent orchestration system for Claude Code.

              ⠀⠀⠀⠀⠀⠀⠀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⡀⠀⠀⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⢠⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡇⠀⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡇⠀⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⢸⡿⠿⠿⠿⠿⠿⠿⠿⠿⠿⠿⠿⠿⠿⠿⢿⣧⠀⠀⠀⠀⠀
              ⢀⣀⣀⣀⣀⣸⣇⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣀⣸⣿⣀⣀⣀⣀⠀
              ⠸⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⠇
              ⠀⠀⠀⠉⢙⣿⡿⠿⠿⠿⠿⠿⢿⣿⣿⣿⠿⠿⠿⠿⠿⢿⣿⣛⠉⠁⠀⠀
              ⠀⠀⠀⣰⡟⠉⢰⣶⣶⣶⣶⣶⣶⡶⢶⣶⣶⣶⣶⣶⣶⡆⠉⠻⣧⠀⠀⠀
              ⠀⠀⠀⢻⣧⡀⠈⣿⣿⣿⣿⣿⡿⠁⠈⢿⣿⣿⣿⣿⣿⠁⠀⣠⡿⠀⠀⠀
              ⠀⠀⠀⠀⠙⣿⡆⠈⠉⠉⠉⠉⠀⠀⠀⠀⠉⠉⠉⠉⠁⢰⣿⠋⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⣿⡇⠀⠀⠀⣠⣶⣶⣶⣶⣶⣶⣄⠀⠀⠀⢸⣿⠀⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⠸⣷⡀⠀⠀⣿⠛⠉⠉⠉⠉⠛⣿⠀⠀⢀⣾⠇⠀⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⠀⠘⢿⣦⡀⣿⣄⠀⣾⣷⠀⣠⣿⣀⣴⡟⠁⠀⠀⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⠀⠀⠀⠙⠻⣿⣿⣿⣿⣿⣿⣿⣿⠟⠁⠀⠀⠀⠀⠀⠀⠀⠀
              ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠈⠙⠛⠛⠋⠁⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀

         ╔══════════════════════════════════════════════════╗
         ║                    A C E                         ║
         ║          Agent Creates Everything                ║
         ╚══════════════════════════════════════════════════╝

    ┌──────────────────────────────────────────────────────────┐
    │  3 Team Leads      →  Unlimited Mini-Agents              │
    │  Persistent State  →  Zero Context Loss                  │
    │  Atomic Commits    →  Perfect Traceability               │
    └──────────────────────────────────────────────────────────┘

What is Ace?

Ace is an orchestration layer for Claude Code that enables:

  • Parallel Execution: 3 Team Leads (Alpha, Beta, Gamma) work simultaneously
  • Unlimited Mini-Agents: Each Team Lead can spawn sub-agents for complex tasks
  • Persistent State: Never lose context between sessions
  • Atomic Commits: One commit per task with full traceability
  • Zone Ownership: No file conflicts between agents
  • Natural Language: Just describe what you want to build

Installation

Option 1: Clone this repository

git clone https://github.com/YOUR_USERNAME/ace.git my-project
cd my-project

Option 2: Copy to existing project

Copy the .ace folder and CLAUDE.md to your project root:

your-project/
├── .ace/           # Copy this folder
├── CLAUDE.md       # Copy this file
└── ... your code

Quick Start

1. Start Claude Code in your project

claude

2. Activate Ace

Just say:

Ace

Or use any natural language:

"I want to build a dashboard"
"Help me create an API"
"Start a new project"

3. Follow the workflow

/ace:new-project "My App"     # Initialize
/ace:create-roadmap           # Plan phases
/ace:plan-phase 1             # Create tasks
/ace:execute-plan             # Build it!

Commands

Project Lifecycle

Command Description
/ace:new-project Initialize project with PROJECT.md
/ace:create-roadmap Build ROADMAP.md and STATE.md
/ace:map-codebase Analyze existing code (brownfield)
/ace:init Quick-start initialization

Phase Management

Command Description
/ace:plan-phase [N] Create PLAN.md for phase N
/ace:execute-plan Run current plan with parallel agents
/ace:add-phase Append phase to roadmap
/ace:insert-phase [N] Insert urgent phase
/ace:remove-phase [N] Remove phase N

Progress & Status

Command Description
/ace:progress Show status and next steps
/ace:status Quick status check
/ace:show-plan Display current plan
/ace:list-phases List all phases

Session Management

Command Description
/ace:pause-work Save progress and create handoff
/ace:resume-work Restore from last session
/ace:resume-task [id] Resume specific task

Verification

Command Description
/ace:verify-work User acceptance testing
/ace:plan-fix Plan fixes for issues
/ace:consider-issues Review deferred issues

Help

Command Description
/ace:help Show all commands
/ace:config View/edit configuration

Architecture

                    ORCHESTRATOR (you)
                          |
          +---------------+---------------+
          |               |               |
       ALPHA           BETA            GAMMA
       Lead            Lead            Lead
          |               |               |
       mini            mini            mini
      agents          agents          agents

How It Works

  1. You describe what you want to build
  2. Orchestrator creates a plan with max 3 tasks
  3. Team Leads execute tasks in parallel
  4. Mini-agents handle complex subtasks
  5. Git commits are made per-task automatically
  6. State is persisted for session continuity

Project Types

Ace auto-detects and optimizes for:

Type Description Use Case
frontend React, Next, Vue, Svelte UI, dashboards
backend Node, Express, APIs Services, APIs
fullstack Full applications End-to-end apps
automation Scripts, CLI tools DevOps, tooling
agentic AI agents, LLM pipelines Autonomous systems
library npm packages Reusable code
monorepo Multi-package Large-scale projects
custom User-defined Any structure

Configuration

Edit .ace/config.json to customize:

{
  "orchestration": {
    "maxTasksPerPlan": 3,
    "showPlanBeforeExecute": true
  },
  "workflow": {
    "mode": "interactive"  // or "yolo" for autonomous
  },
  "qualityGates": {
    "typescript": { "enabled": true },
    "build": { "enabled": true }
  }
}

Natural Language Examples

You don't need to memorize commands. Just talk naturally:

You Say Ace Does
"I want to build a dashboard" Creates project, gathers requirements
"What's the status?" Shows progress board
"Continue where we left off" Resumes from last session
"Build it" Executes current plan
"Check the work" Runs verification

File Structure

.ace/
├── SKILL.md              # Main orchestration logic
├── config.json           # Configuration
├── project-types.json    # Project type definitions
├── commands/             # 26 command definitions
├── templates/            # Document templates
│   ├── planning/         # PROJECT, ROADMAP, STATE, etc.
│   └── codebase/         # Analysis templates
└── scripts/              # Optional terminal scripts

Requirements

  • Claude Code CLI
  • Git (for atomic commits)
  • Node.js (for quality gates)

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License - see LICENSE file for details.


Built with Claude Code

About

Unified multi-agent orchestration layer for Claude Code with parallel execution, session persistence, and natural language routing.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published