to-user-storieslisted
Install: claude install-skill clnnn/playbook
# To User Stories
Convert whatever material is at hand — the current conversation, pasted notes, a PRD, a release slice issue — into implementation-ready user stories, each written for a **fresh agent**: an implementer with the code in front of them but none of this conversation. Every story passes the checks gate, the user confirms the set, and only then do the stories become tickets wired with blocked-by edges.
## Repo wiring
Two pointers, both written by `/setup-playbook-skills`. A missing one: tell the user to run that skill and hold.
- **`docs/agents/issue-tracker.md`** — reached from the `### Issue tracker` sub-block of `## Agent skills` in `AGENTS.md` or `CLAUDE.md`. Its operation table defines every tracker verb this skill names in **bold** — **publish**, **fetch**, **link A blocked by B** and the rest — and is the authority on both the command to run and the fallback when a tracker cannot express an edge. Read it before the first write.
- **`docs/agents/domain.md`** — the rules for reading this repo's `CONTEXT.md` glossary and its ADRs. Follow it at Ground, so the stories name domain concepts in the glossary's words (check 3) and cite ADR numbers rather than restating decisions (§3).
## Input
Anything supplied with the invocation counts as material already given — text after the skill name, a pasted dump, the conversation so far.
**A slice issue** given as a number, path, or URL is **fetched** from the tracker. It carries everything a breakdown needs, and the