To Tickets

To Tickets

Break a plan, spec, or conversation into a set of tickets — tracer-bullet vertical slices, each declaring the tickets that block it.

The issue tracker and triage label vocabulary should have been provided to you — run /setup-matt-pocock-skills if not.

Process

1. Gather context

Work from whatever is already in the conversation context. If the user passes a reference (a spec path, an issue number or URL) as an argument, fetch it and read its full body and comments.

If the source issue already has sub-issues, it is already sliced — stop and list its children instead of slicing again.

2. Explore the codebase (optional)

If you have not already explored the codebase, do so to understand the current state of the code. Ticket titles and descriptions should use the project's domain glossary vocabulary, and respect ADRs in the area you're touching.

Look for opportunities to prefactor the code to make the implementation easier. "Make the change easy, then make the easy change."

3. Draft vertical slices

Break the work into tracer bullet tickets.

• Each slice cuts a narrow but COMPLETE path through every layer (schema, API, UI, tests) — vertical, NOT a horizontal slice of one layer
• A completed slice is demoable or verifiable on its own
• Each slice is sized to fit in a single fresh context window
• Any prefactoring should be done first

Give each ticket its blocking edges — the other tickets that must complete before it can start. A ticket with no blockers can start immediately.

Give each ticket its seams under test — the public boundaries /implement will write its failing tests against (see the tdd skill for what a seam is). Name the interface, not the internals. Agreeing them here is what makes them the "pre-agreed seams" /implement needs, so implementation never starts without a place to put the red test.

Wide refactors are the exception to vertical slicing. A wide refactor is one mechanical change — rename a column, retype a shared symbol — whose blast radius fans across the whole codebase, so a single edit breaks thousands of call sites at once and no vertical slice can land green. Don't force it into a tracer bullet; sequence it as expand–contract. First expand: add the new form beside the old so nothing breaks. Then migrate the call sites over in batches sized by blast radius (per package, per directory), each batch its own ticket blocked by the expand, keeping CI green batch to batch because the old form still exists. Finally contract: delete the old form once no caller remains, in a ticket blocked by every migrate batch. When even the batches can't stay green alone, keep the sequence but let them share an integration branch that all block a final integrate-and-verify ticket — green is promised only there.

4. Quiz the user

Present the proposed breakdown as a numbered list. For each ticket, show:

• Title: short descriptive name
• Blocked by: which other tickets (if any) must complete first
• What it delivers: the end-to-end behaviour this ticket makes work
• Seams under test: the public boundaries its tests will be written against

Ask the user:

• Does the granularity feel right? (too coarse / too fine)
• Are the blocking edges correct — does each ticket only depend on tickets that genuinely gate it?
• Are the seams the right ones to test at?
• Should any tickets be merged or split further?

Iterate until the user approves the breakdown.

5. Publish the tickets to the configured tracker

Publish the approved tickets. How depends on the tracker /setup-matt-pocock-skills configured — the tickets are the same either way, only the shape of the blocking edges changes:

• Local files → write one file per ticket under .scratch/<feature-slug>/issues/<NN>-<slug>.md, numbered from 01 in dependency order (blockers first). Each file's "Blocked by" lists the numbers/titles it depends on. Use the per-ticket file template below — one ticket per file, never a single combined file.
• A real issue tracker (GitHub, Linear, …) → publish one issue per ticket in dependency order (blockers first) so each ticket's blocking edges can reference real identifiers. Set the platform's native blocking / sub-issue relationship where it has one (on GitHub, gh issue edit <n> --add-blocked-by <#>) — and write the ## Blocked by section in addition to it, never instead of it. When the source is a parent issue, link every published ticket as a child using the issue tracker doc's parent-linking convention — then verify the parent reports the full ticket count as children before finishing.

On a tracker, never omit the ## Blocked by section — not on a tracker with native edges, and not on a ticket with no blockers, where it reads None — can start immediately. A ticket with neither native edges nor that section reads as unresolved, not unblocked, and is never dispatched: the frontier reader treats silence as its own answer rather than guessing. Write the section in the grammar that reader parses — the heading exactly Blocked by, one bare #NNN per blocking issue in this repo, or the literal None. The grammar and what each answer means: ~/.agents/skills/burndown/references/frontier.md.

Either way, give each ticket exactly one of three triage roles — its label on a tracker, its Status line in a local file: ready-for-agent when it is fully specified and an agent can build it unattended; ready-for-human when it is specified but a human builds it (a taste call, credentials, something the user wants to write themselves); needs-info when it needs grilling first. A ticket genuinely too unclear to call is needs-info, with its body saying what is unclear. Never needs-triage — that role is for issues outside people open. Why: a ticket marked ready is dispatched as it stands, so one that still needs grilling must carry needs-info or it gets built before anyone grills it.

Work the frontier: any ticket whose blockers are all done. For a purely linear chain that means top to bottom.

When the source was a tracker issue — a spec or plan these tickets now decompose — keep its ready-for-agent label and add spec. The spec label marks the issue as a parent reference for a burn to hand to /implement-spec; plain implement-dispatch refuses a spec-labelled issue, so the label does not send an ordinary agent to build it unsliced. Leave the issue open as the tickets' parent reference; don't otherwise edit its contents.

—

TL;DR: one or two sentences of plain English — what this ticket gets working, said so anyone (non-engineer included) understands it. No domain jargon, no coined terms, no glossary words that need a gloss. Everything below this line may use the project's full technical vocabulary freely.

What to build: the end-to-end behaviour this ticket makes work, from the user's perspective — not a layer-by-layer implementation list.

Blocked by: the numbers/titles of the tickets that gate this one, or "None — can start immediately".

Seams under test: the public boundaries the tests go against — user-approved, so /implement writes its failing tests here without asking again.

Status: the triage role step 5 chose — ready-for-agent, ready-for-human, or needs-info

• Acceptance criterion 1
• Acceptance criterion 2

Parent

A reference to the parent issue on the tracker (if the source was an existing issue, otherwise omit this section).

TL;DR

One or two sentences of plain English — what this ticket gets working, said so anyone (non-engineer included) understands it. No domain jargon, no coined terms, no glossary words that need a gloss. Every section below may use the project's full technical vocabulary freely.

What to build

The end-to-end behaviour this ticket makes work, from the user's perspective — not layer-by-layer implementation.

Acceptance criteria

• Criterion 1
• Criterion 2

Blocked by

• A reference to each blocking ticket, or "None — can start immediately".

Seams under test

• The public boundaries the tests go against — user-approved, so /implement writes its failing tests here without asking again.

In either form, avoid specific file paths or code snippets — they go stale fast. Exception: if a prototype produced a snippet that encodes a decision more precisely than prose can (state machine, reducer, schema, type shape), inline it and note briefly that it came from a prototype. Trim to the decision-rich parts — not a working demo, just the important bits.

Publishing local ticket files, and amending them later, both change tracked files in the working checkout. Commit and push them to main before the next /implement — that command branches from the pushed main, so an uncommitted ticket edit is either missing inside the worktree or duplicated into it, and a duplicate that then advances on the branch conflicts when the work ships.

Work the frontier one ticket at a time with /implement, clearing context between tickets.