<aside> 📌 The day-by-day D-POAF® loop. Pin this page for team reference.

</aside>

TL;DR — the loop in 60 seconds

Open Wave Scope → For each AI generation, open Prompt Action → log model + config + context → invoke AI → update PA with output → commit with [AI:model:PA-ID] → open PR → close with Proof Record

Day 1 — Open a Wave

Add a row to the 🌊 Waves database. Fill Wave ID, Name, Profile, Captain, Business Objective, AI Tools Authorized, Scope, and PoD/PoV/PoR criteria. The row IS your Wave Scope Document.

Days 2-N — Log every Prompt Action

Per Dynamic Law DL-006: log EVERY PA before invoking the AI. Fill PA-ID, Wave (relation), Sub-Phase, Role, Model identity, Model config, Context source, Prompt. Update Quality + Output after generation.

Day N — Close the Wave

Add a row to ✅ Proof Records. Fill PoD/PoV/PoR with evidence and sign-offs. Update Wave status to Closed.

6 months later — Trace back

Need to know why an artifact was built? Find the PA-ID in commit message → look up the PA row → see model + config + context → walk back to Wave row → see business intent. Total time: 5-10 minutes.