mp-to-issues

---
name: mp-to-issues
description: Use when breaking a plan, spec, or PRD into independently-grabbable tracker issues — vertical tracer-bullet slices that cut through every layer end-to-end, each one demoable on its own.
source: https://github.com/mattpocock/skills/blob/main/skills/engineering/to-issues/SKILL.md
generated: 2026-05-12T18:05:22.410Z
category: concept
audience: engineers
---

## Tutorials

- https://skillmake.xyz/v/mp-to-issues.mp4

## When to use

- Converting a PRD or design doc into implementation tickets ready for AFK agents
- Splitting a large feature into thin slices a single contributor (human or agent) can finish in one sitting
- Marking which slices are HITL (human interaction required) vs AFK (agent-mergeable)
- Sequencing tickets so blockers ship first and downstream slices can reference real issue IDs

## Key concepts

### vertical tracer-bullet slice

Each issue is a thin vertical slice that cuts through ALL integration layers — schema, API, UI, tests — end-to-end. Not a horizontal slice of one layer. A completed slice is demoable or verifiable on its own. Prefer many thin slices over few thick ones.

### HITL vs AFK

Slices are tagged HITL (human-in-the-loop — requires architectural decision, design review, or external access) or AFK (agent-mergeable without human interaction). Prefer AFK over HITL where possible — that's where the leverage is.

### quiz the user before publishing

Present the proposed breakdown as a numbered list with title, type (HITL/AFK), blocked-by, and user stories covered. Ask whether granularity feels right (too coarse/too fine), whether dependencies are correct, whether any slices should merge or split, whether HITL/AFK tags are right. Iterate until approved.

### publish in dependency order

Publish issues in dependency order — blockers first — so you can reference real issue IDs in the 'Blocked by' field of downstream slices. Skipping this leaves dangling references.

### respect domain glossary

Issue titles and descriptions must use the project's CONTEXT.md vocabulary and respect ADRs in the area you're touching. Drift from canonical terms means downstream contributors won't trust the issue.

### no file paths or code in issues

Issue bodies describe end-to-end behavior, not layer-by-layer implementation. Avoid specific file paths or code snippets — they go stale fast. Exception: a prototype-derived state machine / reducer / schema / type shape can be inlined when it encodes a decision more precisely than prose. Trim to decision-rich parts, not a working demo.

### do NOT modify the parent issue

When breaking a parent issue into slices, do not close or modify the parent. The parent stays as the tracking thread; child issues link back via the 'Parent' field of the template.

## API reference

```
Vertical-slice rules
```

The shape every slice must have. Failing these turns 'issues' into a layer-by-layer to-do list that can't be parallelised.

```
- Each slice delivers a narrow but COMPLETE path through every layer (schema, API, UI, tests)
- A completed slice is demoable or verifiable on its own
- Prefer many thin slices over few thick ones
```

```
Per-slice review summary (present to user)
```

What you show the user for each proposed slice when quizzing for approval. The user signs off before any issue is published.

```
- Title: short descriptive name
- Type: HITL / AFK
- Blocked by: which other slices (if any) must complete first
- User stories covered: which user stories this addresses (if the source has them)

Ask:
- Does the granularity feel right? (too coarse / too fine)
- Are the dependency relationships correct?
- Should any slices be merged or split further?
- Are the correct slices marked as HITL and AFK?
```

```
Issue body template
```

Use this template verbatim for every published issue. Apply the ready-for-agent triage label unless instructed otherwise.

```
## Parent

A reference to the parent issue (omit if no parent).

## What to build

A concise description of this vertical slice. Describe the end-to-end behavior, not layer-by-layer implementation. Avoid file paths or code snippets — they go stale. Inline a prototype-derived snippet only when it encodes a decision more precisely than prose.

## Acceptance criteria

- [ ] Criterion 1
- [ ] Criterion 2
- [ ] Criterion 3

## Blocked by

- A reference to the blocking ticket (if any)

Or "None - can start immediately" if no blockers.
```

## Gotchas

- Each slice must cut through EVERY layer. A schema-only or UI-only ticket is a horizontal slice and can't be demoed standalone.
- Prefer AFK over HITL where possible — HITL slices block on humans and don't parallelise.
- Publish in dependency order so 'Blocked by' fields can reference real issue IDs, not 'TBD'.
- Use the project's CONTEXT.md vocabulary in titles and descriptions — don't invent new terminology.
- Avoid specific file paths or code snippets in issue bodies — they rot fast.
- Exception: inline a prototype-derived state machine / schema / type shape when it encodes a decision more precisely than prose, trimmed to the decision-rich parts.
- Do NOT close or modify the parent issue when splitting it into slices — the parent stays as the tracking thread.
- Quiz the user on granularity, dependencies, and HITL/AFK tags BEFORE publishing — once the issues are live they're expensive to restructure.

---
Generated by SkillMake from https://github.com/mattpocock/skills/blob/main/skills/engineering/to-issues/SKILL.md on 2026-05-12T18:05:22.410Z.
Verify against source before relying on details.

File: ~/.claude/skills/mp-to-issues/SKILL.md