Claude Orchestrator AI Evolution

Claude Orchestrator AI Evolution

[](https://github.com/eroslifestyle/Claude-Orchestrator-AI-Evolution) [](LICENSE) [](https://www.python.org/) [](https://github.com/eroslifestyle/Claude-Orchestrator-AI-Evolution/actions) [](https://github.com/eroslifestyle/Claude-Orchestrator-AI-Evolution)

---

Overview

Claude Orchestrator AI Evolution is a self-optimizing multi-agent orchestration system for Claude Code. It provides intelligent task decomposition, predictive agent caching, adaptive token budgeting, and a modular plugin architecture.

The orchestrator automatically routes tasks to specialized agents based on keyword patterns, complexity scoring, and ML-based pattern recognition with >90% accuracy.

Key Stats

| Metric | Value | |--------|-------| | Agents | 43 (6 Core + 22 L1 + 15 L2) | | Skills | 30 | | Languages Supported | 6 | | Tests | 400+ | | Lines of Code | 64K+ |

---

Features

43 Specialized Agents

| Tier | Count | Examples | |------|-------|----------| | Core (L0) | 6 | Orchestrator, Analyzer, Coder, Reviewer, Documenter, System Coordinator | | L1 Experts | 22 | GUI Super Expert, Database Expert, Security Expert, API Expert, DevOps Expert, MQL Expert, Mobile Expert, N8N Expert, etc. | | L2 Specialists | 15 | Layout Specialist, Query Optimizer, Auth Specialist, Endpoint Builder, Unit Test Specialist, MQL Optimization, etc. |

30 Skills

| Category | Skills | |----------|--------| | Core | orchestrator, code-review, git-workflow, testing-strategy, debugging, api-design, remotion-best-practices, keybindings-help | | Utility | strategic-compact, verification-loop, checkpoint, status, metrics, prompt-engineering-patterns | | Workflow | plan, tdd-workflow, security-scan, refactor-clean, build-fix, multi-plan, fix, cleanup, simplify | | Language | python-patterns, python-performance-optimization, typescript-patterns, go-patterns | | Learning | learn, evolve |

AI-Native Capabilities

• Predictive Agent Caching - ML-based pattern recognition (>90% accuracy)
• Adaptive Token Budgeting - Dynamic 200-1500 token allocation
• A/B Testing Framework - Multi-variant experiments with statistical significance
• Auto-Tuning Parameters - Bayesian optimization for system parameters
• Chaos Engineering - 5 failure types injection for resilience testing
• Distributed Locking - Redis-based coordination

---

Quick Start

1. Clone Repository

git clone https://github.com/eroslifestyle/Claude-Orchestrator-AI-Evolution.git
cd Claude-Orchestrator-AI-Evolution

2. Install Dependencies

pip install -r requirements.txt

3. Run Tests

# Unit tests with coverage
pytest lib/tests/ -v --cov=lib --cov-report=term-missing

# Or use the provided scripts
./scripts/run_coverage.sh      # Unix/Linux/Mac
.\scripts\run_coverage.ps1     # Windows

4. Use with Claude Code

Copy the .claude/ directory to your project or set as global:

# Windows
xcopy /E /I .claude %USERPROFILE%\.claude

# Linux/Mac
cp -r .claude ~/.claude

---

Architecture

Project Structure

.claude/
├── agents/              # 43 agent definitions
│   ├── core/            # L0: 6 core agents
│   ├── experts/         # L1: 22 expert agents
│   │   └── L2/          # L2: 15 specialist agents
│   ├── config/          # Routing, standards
│   └── system/          # Registry, protocols
├── skills/              # 30 skills + orchestrator/
│   └── orchestrator/    # Main orchestrator skill
│       ├── SKILL.md     # Skill definition
│       ├── VERSION.json # Version tracking
│       └── docs/        # 18+ documentation files
├── lib/                 # Python modules (91 files, 64K+ LOC)
│   ├── predictive_cache.py
│   ├── adaptive_budget.py
│   ├── ab_testing.py
│   ├── auto_tuner.py
│   ├── facade.py
│   └── ... (87 more)
├── rules/               # 11 rules files
│   ├── common/          # Universal rules
│   ├── python/          # Python patterns
│   ├── typescript/      # TypeScript patterns
│   ├── go/              # Go patterns
│   ├── rust/            # Rust patterns
│   ├── java/            # Java patterns
│   └── cpp/             # C++ patterns
├── templates/           # 3 templates
└── workflows/           # 4 workflows

Documentation

| Document | Description | |----------|-------------| | [architecture.md](skills/orchestrator/docs/architecture.md) | System architecture | | [algorithm.md](skills/orchestrator/docs/algorithm.md) | Step-by-step algorithm | | [routing.md](skills/orchestrator/docs/routing.md) | Complete routing table | | [agents.md](skills/orchestrator/docs/agents.md) | Agent hierarchy (L0/L1/L2) | | [setup-guide.md](skills/orchestrator/docs/setup-guide.md) | Setup guide | | [examples.md](skills/orchestrator/docs/examples.md) | Usage examples | | [test-suite.md](skills/orchestrator/docs/test-suite.md) | Test documentation |

---

Configuration

Environment Variables

| Variable | Description | Default | |----------|-------------|---------| | SILENT_START | Suppress initial task table | false | | OUTPUT_MODE | Output verbosity (verbose/compact/silent) | compact |

Settings File

{
  "env": {
    "SILENT_START": "false",
    "OUTPUT_MODE": "compact"
  }
}

---

Supported Languages

| Language | Rules File | Patterns | |----------|------------|----------| | Python | rules/python/patterns.md | PEP 8, type hints, async, testing | | TypeScript | rules/typescript/patterns.md | strict mode, zod, discriminated unions | | Go | rules/go/patterns.md | error handling, interfaces, concurrency | | Rust | rules/rust/patterns.md | ownership, borrowing, async, error handling | | Java | rules/java/patterns.md | Spring Boot, DI, testing patterns | | C++ | rules/cpp/patterns.md | Modern C++17/20, RAII, memory management |

---

Changelog

[2026-03-10] V18.1.0 - Unified Evolution & Modular Architecture

Added:

• rules/rust/patterns.md: Rust coding standards
• rules/java/patterns.md: Java patterns (Spring Boot, DI, testing)
• rules/cpp/patterns.md: C++ patterns (Modern C++17/20, RAII)
• scripts/run_coverage.sh and run_coverage.ps1: Coverage scripts

Changed:

• SKILL.md modularizzato: Orchestrator ora include docs/ modulari
• VERSION.json: Aggiornato a V18.1.0, orchestrator V12.9.1

Fixed:

• Sincronizzazione completa lib/ (tutti i moduli Python V17)
• Rimossi file obsoleti e duplicati

[2026-02-18] Agent Teams Gap Fix

• Aggiunta sezione PREREQUISITES per Agent Teams
• Aggiunto disclaimer EXPERIMENTAL FEATURE
• Documentate 7 limitazioni note

[2026-02-16] Dual Provider Setup

• Sincronizzazione assets Anthropic -> GLM
• Setup doppia istanza VS Code
• Verifica integrita completa

[2026-02-26] V12.0 - Deep Audit

• Baseline V12 con audit completo
• Anti-patterns documentati
• Error recovery protocol

[2026-02-15] V11.0 - Initial Release

• Orchestrator core system
• 43 agents, 30 skills
• Modular architecture

---

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit changes using conventional commits (git commit -m 'feat: add amazing feature')
4. Push to branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Commit Convention

(): 

Types: feat, fix, refactor, test, docs, chore, perf, ci, build, style

---

License

MIT License - See [LICENSE](LICENSE) for details.

---

Acknowledgments

• Anthropic for Claude API
• All contributors and testers
• The open-source community

---

Made with care by eroslifestyle

V18.1.0 - Unified Evolution - March 2026

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: eroslifestyle
• Source: eroslifestyle/Claude-Orchestrator-AI-Evolution
• License: MIT

Install and usage instructions live in the source repository linked above.