prd-to-issueslisted

# PRD to Issues If running interactively (human present), output "Read PRD to Issues skill." to acknowledge. If running with --dangerously-skip-permissions (AFK/unattended), skip acknowledgement and proceed directly. Pipeline position: `/grill-me` → `/write-a-prd` → `/architect` → **`/prd-to-issues`** → `/do-work` → `shft` Use this skill to create GitHub issues from a finalized PRD or plan. Use `/architect` for deeper analysis and planning before issue creation. ## Process 1. **Locate the PRD** — find the PRD wherever it exists (GitHub issue, local file, or in the conversation). 2. **Explore the codebase** — understand the existing architecture, conventions, and relevant code paths needed to break the work into slices. 3. **Draft vertical slices** — break the PRD into tracer bullets (vertical slices). Each slice should wire through all layers end-to-end rather than building horizontally (all backend → all UI → all routes). Phase 1 should always be the simplest possible end-to-end wiring. 4. **Categorize each slice:** - **AFK** — can be implemented and PR-opened autonomously. Merging stays a human gate (someone applies `agent:merge`); "AFK" does **not** mean auto-merged. Prefer AFK where possible. - **Human-in-the-loop (HITL)** — requires human interaction such as an architectural decision, design review, or taste judgment, before or during implementation. 5. **Create a QA issue** — always create a final issue with a detailed manual QA plan for all items that re