post-analiz-handoff
Use when the stakeholder asks about an analysis outcome or its implementation tasks - understand the human approval gate and review the tasks the architect created once the human approved
- License 许可
- Apache-2.0
- 最近更新
- 2026-11-15
- Stars
- ★ 122
- Forks
- 18
01 前言 Front Matter
- name
- post-analiz-handoff
- category
- pm
- description
- Use when the stakeholder asks about an analysis outcome or its implementation tasks - understand the human approval gate and review the tasks the architect created once the human approved
03 SKILL.md 正文
Post-Analiz Handoff
Overview
The analiz has a human approval gate. The system-architect does NOT create implementation tasks when it finishes analysis — it presents the plan in the analiz_review column, the human approves it, and only then does the architect create the tasks. Your job is to understand this sequence and review the tasks after they appear, not to create them.
The sequence (who does what)
- Architect writes the spec/plan and moves the analiz task to
analiz_review, then stops. - Human reviews the plan in
analiz_reviewand moves it todone(approve) orneed_revision(reject). This is the human's decision, not yours — do not move an analiz task out ofanaliz_review. - Architect (on approval) creates one implementation task per project, lists them in a comment, and moves the analiz task to
released.
Your steps
Trigger: the stakeholder asks what an analysis found, or about the tasks it produced. The board never dispatches you here on its own — the PM is subscribed to no board column, so this skill only runs when a human brings it up in chat.
list_board_tasksto find the implementation tasks the architect created.- Read the architect's spec/plan with
list_task_documentson the ANALIZ task — they are documents attached to it, never committed to the repository — and read the task comments. - Check each created task carries
derived_frompointing at the analiz task.update_board_taskcannot setderived_from— it has no such field and silently ignores it. If the task has not started (backlog/todo/blocked),delete_board_task+create_board_taskit withderived_fromset, keeping its title, fields and criteria. If work has already started,add_task_comment: "Spec and plan:list_task_documents A-N" — so the developer can still reach them without touching a task mid-flight. - Check the order is declared, not just described. A frontend task that consumes a new endpoint should carry
blocked_byanddeploy_depends_onnaming the backend task; a "Depends on: …" sentence in the description enforces nothing.update_board_taskCAN add toblocked_by/deploy_depends_on(it adds to whatever is already there) — use it for tasks still in backlog/todo/blocked. - If stakeholder decisions were logged as questions during analysis, confirm answers exist.
- Review scope and priority. Edit acceptance criteria with
update_board_taskonly on tasks still inbacklog/todo/blocked— it replaces the whole array and resets every tick. On a task that has started, usecancel_criterionto drop one, or open a new task for added scope (see scope-management); never duplicate or re-create the architect's tasks. - Tell the stakeholder which implementation tasks are queued, in what order, and what happens next. A task sitting in
blockedbecause its blocker is still open is on schedule, not stuck — it is picked up automatically.
Common Mistakes
- Creating implementation tasks yourself — the architect owns decomposition, gated on human approval.
- Moving an analiz task from
analiz_reviewtodone— that is the human's approval action. - Expecting the tasks to exist as soon as the analiz hits
done— they appear after the architect releases the analiz. - Trying to fix a missing
derived_fromwithupdate_board_task— it silently ignores the field; delete and recreate instead. - Editing acceptance criteria on a task already in
in_progressor later — it resets every tick the developer made.
Red Flags
- You are about to
create_board_taskfor work the architect analyzed → stop; the architect creates it after approval. - You moved an analiz task out of
analiz_review→ that is the human's decision. - You called
update_board_task(derived_from=...)expecting it to take effect → it won't; delete + recreate instead.
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