new-ticket

name	new-ticket
description	Use when starting a new piece of work — "/new-ticket", "start a new ticket", "let's start on ticket 1234". Runs branch check → short interview → hands off to Plan Mode for the SPEC. Pairs with the `commit-push` and `open-pr` skills for the rest of the cycle.

Starting a new ticket

1. Branch check

• git status must be clean before switching branches — if not, stop and flag it rather than discarding anything.
• Fetch and pull the latest main. If not currently on main, confirm before switching.

2. Interview

Ask two short questions, one at a time (see CLAUDE.md's Output style):

1. Trello card number?
2. One-line description of the work?

Don't try to gather full requirements here — that's what Plan Mode does next.

3. Derive the branch name

<trello-number>-<short-description>, slug hyphenated and truncated to fit the repo's 16-character limit (see "Environment & workflow" in CLAUDE.md). Create and check out the branch from latest main.

4. Hand off to Plan Mode for the SPEC

Call EnterPlanMode and let Claude Code's normal phased workflow (Explore → Design → Review → Final Plan → ExitPlanMode) do the deeper requirements interview and produce the plan. Don't build a separate SPEC format — the plan file is the SPEC.

5. After the plan is approved

Implement, then use commit-push (repeatedly, per logical change) and open-pr (once, at the end) to finish the cycle. There's no automatic memory save at any point — only save something to memory if it looks genuinely useful, and ask first.