pdlc-skills

Turn AI software engineering into an auditable, on-disk state machine. A staged PDLC workflow (PRD, design, TDD, implement, review, ship, retro) enforces hard contracts — artifacts on disk, per-feature state machine, tests-before-code, objective checks from real command exit codes, single-shot auto-repair — so AI work produces real, reviewable files instead of chat-only output. Because the checks are objective, an autonomous loop can drive tdd→implement→review to convergence unattended (fail-stop / budget guards); because the state machine lives in your repo, it's tool-agnostic — richest integration in Claude Code, plus Codex and others via adapters.