scope-management
Use when defining a feature - name what is in scope, out of scope, and assumed, and handle mid-flight expansion without corrupting in-progress tasks
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-15
- Stars
- ★ 122
- Forks
- 18
01 前言 Front Matter
- name
- scope-management
- category
- pm
- description
- Use when defining a feature - name what is in scope, out of scope, and assumed, and handle mid-flight expansion without corrupting in-progress tasks
03 SKILL.md 正文
Scope Management
Overview
Scope is defended by what you write DOWN, especially the "out of scope" line. Undocumented scope drifts; a task whose AC quietly grows mid-flight is the classic cause of a blown estimate and a bounced UAT.
Core principle: Name the boundary explicitly. New scope becomes a new task, never an edit to an in-progress one.
For every feature, write
- In scope: concrete deliverables in this iteration.
- Out of scope: explicitly named items deferred or excluded — this line does the real work.
- Assumptions: constraints treated as given.
TaskTrooper has no epic. For a single direct task, this lives in its own description. For a feature that spans several tasks via analiz, it lives on the analiz task (description or add_task_document) — never re-written per implementation task.
Mid-flight expansion
When a request grows while work is in progress: create a new task for the expansion. Do NOT edit the AC of an in-progress task — a moving target invalidates the developer's plan and QA's scenarios and silently inflates the estimate.
Except when the human changes the task themselves. A comment the task's owner writes ON the task ("also show the logos in the marquee"), or their own edit of its description or criteria, is not mid-flight expansion to push back on — it is the task's requirement now. It outranks the original description and its out-of-scope list; every role builds, reviews, tests and accepts against it. Do not split it into a new task or treat it as creep unless the human asks for that.
MoSCoW inside a feature
Must → this task's acceptance_criteria. Should/Could → separate backlog tasks, never folded into this one. Won't → the Out of scope line, with a reason (see backlog-prioritization).
Dropping a criterion deliberately
Use cancel_criterion with a reason when a criterion is descoped, superseded by a later human comment, or moved to another task — never to get past work that simply wasn't done. cancel_criterion says "we're not shipping this here," which is a scope decision; a rejected review_criterion says "this doesn't work yet," which is a defect. Don't use one for the other.
A new idea found during UAT
Something the PM or QA notices in QA or human_uat that nobody asked for is a candidate for a new backlog task (create_board_task, priority low, same repository/project) — it is never a gap on the current task and never blocks its approval. Only something that breaks a criterion, a human comment, or the original request is a gap.
Worked Example
Feature: "task export."
- In scope: CSV export of a project's tasks via the board UI.
- Out of scope: PDF export, scheduled/emailed exports, custom column selection.
- Assumptions: export is on-demand; ≤ 10k tasks per project.
Mid-sprint the stakeholder says "also email it weekly." That's the out-of-scope "scheduled export" — create a new task for it, keep the in-progress CSV task untouched, and tell the stakeholder it's queued separately.
Common Mistakes
- Omitting the "out of scope" line → scope creeps by default.
- Editing an in-progress task's AC to absorb new asks.
- "We'll figure out scope later" → later is a UAT surprise.
Red Flags
- A feature with an in-scope list but no out-of-scope list.
- An in-progress task whose AC changed after work started.
04 同仓库更多 Skills
visual-craft
暂无简介
ux-copy
Use when you write any words a user reads in a mockup or a hand-off - content language and voice, buttons, labels, errors, empty states, confirmations, numbers and dates, and the longest-string check
screen-mockup-html
Use when you draw screens as the design task's HTML mockup - ONE canvas document holding every screen and every variant side by side, the sanitizer-safe recipe, phone and wide frames, the states grid, static control states, and the loopback self-check with screenshots at 375, 768 and 1440 attached to the task
performance-awareness
How the performance score moves for your role. Use when the score context shows a declining trend, or before handing off work you have not self-checked against every criterion.
design-variants
Use when a screen design needs variants - how many, variants as side-by-side rows on the one canvas, what makes two variants genuinely different, recommending one, how the human's choice is read, and narrowing to the chosen one in revision
design-system-authoring
Use when you derive, change or extend a design system - where tokens live per stack, the DTCG token tree, role-based names for both themes, base vs layer vs drift, the DESIGN.md section order, the inventory format, the contrast check and the lint every proposal returns