aidlc-skill_skill

Overview

This skill orchestrates the AI-Driven Development Life Cycle (AI-DLC) to structure software projects across Inception, Construction, and Operations phases. It guides planning, requirements, design, code generation, and build/test activities while enforcing human approvals and a complete audit trail. Use it to run adaptive, repeatable development workflows tailored to project complexity.

How this skill works

On activation it loads required reference material and initializes a project state and audit log. The workflow proceeds through Inception (workspace detection, requirements, planning, optional reverse engineering and design), Construction (design variants, code generation, build and test), and an extendable Operations phase. Each stage follows a two-part pattern: Planning (plan, multiple-choice [Answer]: questions, collect answers, require approval) and Generation (execute approved plan, generate artifacts, log actions, require approval).

When to use it

• When a request begins with "Using AI-DLC" or asks for structured AI-driven development
• For projects that need requirements analysis, architecture, and staged implementation
• When you want an auditable, human-in-the-loop workflow with explicit approvals
• For incremental or adaptive delivery where stage depth should vary by complexity
• When you need reproducible plans, artifacts, and build-and-test automation

Best practices

• Always load the prescribed reference documents at workflow start to ensure compliance
• Keep user interactions in multiple-choice [Answer]: format to avoid ambiguity
• Require explicit user approval before moving between planning and generation steps
• Log every interaction with ISO 8601 timestamps and store raw user input in the audit
• Place application code in the workspace root and keep workflow artifacts in the aidlc docs area

Example use cases

• Kick off a new product: define requirements, generate user stories, design modules, then scaffold code
• Onboard to an existing codebase: run workspace detection and conditional reverse engineering to produce requirements and plans
• Iterative feature delivery: plan a minimal depth run, generate code for a unit, run build-and-test, collect approval, repeat
• Create compliance-friendly artifacts: generate audit logs and state documents for process traceability