GitHub Template AI Agents

Production-ready template for AI agent-powered development with Claude Code, Gemini CLI, OpenCode, Qwen Code, and more.

Quick Links: Quick Start · Documentation · Contributing

What Is This?

A unified harness for AI coding agents that provides consistent workflows across Claude Code, Gemini CLI, OpenCode, Qwen Code, Windsurf, Cursor, and Copilot Chat. Built for teams who want to scale AI-assisted development with quality gates, skills, and sub-agent patterns.

Key Features:

✓ Multi-Agent Support: Works with 6+ AI coding tools simultaneously
✓ Skills System: Reusable knowledge modules in canonical location
✓ Quality Gates: Automatic lint, test, format before commits
✓ Context Discipline: Prevents context rot with sub-agents and hooks
✓ Dependabot Integration: Automated security and version updates

Quick Start (2 Minutes)

Prerequisites

One or more AI coding CLI tools (Claude Code, Gemini CLI, OpenCode, Qwen Code)
Git 2.30+ (install)

Installation

# Use this template on GitHub, then clone
git clone https://github.com/your-org/your-project.git
cd your-project

Setup

# Create skill symlinks (run once)
./scripts/setup-skills.sh

# Install git pre-commit hook
cp scripts/pre-commit-hook.sh .git/hooks/pre-commit && chmod +x .git/hooks/pre-commit

Verify

./scripts/validate-skills.sh

Expected output:

✓ All skill symlinks intact
✓ SKILL.md files valid

Core Concepts

Single Source of Truth

All agents read from AGENTS.md - CLI-specific files (.claude.md, .gemini.md) contain only overrides.

AGENTS.md → Single source of truth
├── CLAUDE.md → Overrides only (@AGENTS.md)
├── GEMINI.md → Overrides only (@AGENTS.md)
└── opencode.json → Configuration

Skills with Progressive Disclosure

Skills live canonically in .agents/skills/. Claude Code and Qwen Code use symlinks; Gemini CLI and OpenCode read directly from .agents/skills/:

.agents/skills/           # Canonical source (single location)
├── task-decomposition/
├── shell-script-quality/
└── github-readme/

.claude/skills/           # Symlinks → ../../.agents/skills/
.qwen/skills/             # Symlinks → ../../.agents/skills/

Sub-Agent Patterns

Delegate isolated tasks to sub-agents for context isolation:

graph LR
    A[Main Agent] --> B[Sub-Agent 1]
    A --> C[Sub-Agent 2]
    B --> D[Task Complete]
    C --> E[Task Complete]
    D --> F[Synthesize]
    E --> F

Usage Examples

Basic: Run with Claude Code

claude "Implement user authentication with OAuth2"

Advanced: Multi-Agent Coordination

# Main agent decomposes task, delegates to sub-agents
claude "/task-decomposition Implement the new API endpoint"

Quality Gate (Automatic)

# Pre-commit hook runs automatically
git commit -m "feat: add user registration"

# Or run manually
./scripts/quality_gate.sh

Documentation

📚 AGENTS.md - Main agent instructions (single source of truth)
📖 Quick Start - 5-minute setup guide
🏗️ Harness Overview - Architecture and patterns
🪝 Skills Guide - Creating reusable skills
🤖 Sub-Agents - Context isolation patterns
🔧 Hooks - Pre/post tool hooks
📊 Context - Back-pressure mechanisms
🔄 Migration - Adopting in existing projects

Available Skills

Skill	Description
`task-decomposition`	Break complex tasks into atomic goals
`shell-script-quality`	Lint/test shell scripts (ShellCheck + BATS)
`github-readme`	Create human-focused README.md files
`parallel-execution`	Coordinate parallel agent execution
`iterative-refinement`	Progressive improvement loops
`agent-coordination`	Multi-agent orchestration patterns
`goap-agent`	Goal-oriented action planning
`web-search-researcher`	Web research and synthesis
`eu-ai-act-compliance`	⚙️ EU AI Act regulatory adherence
`durable-objects`	⚙️ Cloudflare Durable Objects coordination

Contributing

We welcome contributions! See our Contributing Guide for:

Development environment setup
Good first issues
Code style and testing requirements
Pull request process

Community

🐛 Issue Tracker
- Report bugs
📝 Discussions
- Ask questions

License

This project is licensed under the MIT License - see the LICENSE file for details.

Built with AI agents. Maintained by humans.

Name	task-decomposition
Description	Break down complex tasks into atomic, actionable goals with clear dependencies and success criteria. Use this skill when you need to plan multi-step projects, coordinate agents, or decompose complex user requests into manageable sub-tasks.