Skip to content

konippi/aidlc-skill

aidlc-skill

License: MIT Agent Skills

AI-Driven Development Life Cycle (AI-DLC) as an Agent Skill for any compatible coding agent.

What is AI-DLC

AI-DLC is an AI-centric transformative approach to software development that emphasizes two powerful dimensions:

  • AI Powered Execution with Human Oversight: AI systematically creates detailed work plans, actively seeks clarification, and defers critical decisions to humans.
  • Dynamic Team Collaboration: AI handles routine tasks so teams can focus on problem solving, creative thinking, and rapid decision-making.

Learn more: the AI-DLC blog post and the Method Definition Paper.

Install

npx skills add konippi/aidlc-skill

Or manually (pick the path your agent uses):

git clone https://github.com/konippi/aidlc-skill.git
ln -s /path/to/aidlc-skill ~/.claude/skills/aidlc-skill
ln -s /path/to/aidlc-skill ~/.kiro/skills/aidlc-skill
ln -s /path/to/aidlc-skill ~/.agents/skills/aidlc-skill

For project-scoped installation, symlink into .claude/skills/, .kiro/skills/, or .agents/skills/ in your project root instead.

Note

Kiro CLI users: After installing skills, add them to your custom agent's resources in .kiro/agents/<agent>.json:

{
  "resources": ["skill://.kiro/skills/**/SKILL.md"]
}

Usage

Start any software development request with:

Using AI-DLC, build a REST API for user management

The agent displays a welcome message, detects whether the workspace is greenfield or brownfield, and walks through the three-phase adaptive workflow with explicit approval gates between stages.

Three-phase lifecycle

Phase Purpose Key stages
Inception What to build and why Workspace Detection, Reverse Engineering, Requirements Analysis, User Stories, Workflow Planning, Application Design, Units Generation
Construction How to build it (per-unit loop + tests) Functional Design, NFR Requirements, NFR Design, Infrastructure Design, Code Generation, Build and Test
Operations Placeholder for future deploy/monitor Operations (placeholder)

See SKILL.md for the exact activation contract, or open any references/<phase>/<stage>.md for stage-level detail.

Generated artifacts layout

Workflow output lands in aidlc-docs/ in your workspace. Application code lands in the workspace root — never in aidlc-docs/:

<workspace>/
├── [your application code]
└── aidlc-docs/
    ├── aidlc-state.md
    ├── audit.md
    ├── inception/
    ├── construction/{unit}/
    └── operations/

Contributing

See CONTRIBUTING.md.

Security

See SECURITY.md for reporting vulnerabilities.

License

MIT

About

AI-Driven Development Life Cycle (AI-DLC) as Agent Skills for AI coding agents.

Topics

Resources

License

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Contributors