返回技能列表

tool-design-sprint-sketch

product-on-purpose
更新于 2 days ago
7 次查看
238
33
238
在 GitHub 上查看
design

关于

This skill orchestrates the second day (Tuesday) of a Design Sprint, guiding teams through lightning demos and the structured four-step sketching protocol (Notes, Ideas, Crazy 8s, Solution Sketch). It facilitates the process for each member to create one independent solution sketch but does not generate the sketches itself. The output organizes the demo board, sketch assignments, and the collected sketches for the next day's review.

快速安装

Claude Code

推荐
主要方式
npx skills add product-on-purpose/pm-skills -a claude-code
插件命令备选方式
/plugin add https://github.com/product-on-purpose/pm-skills
Git 克隆备选方式
git clone https://github.com/product-on-purpose/pm-skills.git ~/.claude/skills/tool-design-sprint-sketch

在 Claude Code 中复制并粘贴此命令以安装该技能

技能文档

<!-- PM-Skills | https://github.com/product-on-purpose/pm-skills | Apache 2.0 -->

Design Sprint Sketch (Tuesday)

Structure Tuesday's solo-but-together work. Each team member, working independently and silently, produces lightning demos in the morning and a four-step solution sketch in the afternoon. The skill structures the activity; the humans produce the sketches. Wednesday's heat-map orientation depends on having a cohort of independent sketches that did NOT contaminate each other through group brainstorming.

Family contract: docs/reference/skill-families/design-sprint-skills-contract.md. This skill is a member of design-sprint-skills.

When to Use

  • It is Day 2 of the Design Sprint and Monday's Map and Target artifact is signed off.
  • Each team member brings at least 3 lightning demo sources (existing products, references, analogies that bear on the challenge).
  • The team has accepted the silent-independent-sketching constraint and will NOT lapse into group brainstorming.
  • Sketches are due Tuesday end-of-day for Wednesday morning heat-map.

When NOT to Use

  • Monday is not closed. Return to tool-design-sprint-map-and-target.
  • The team's instinct is to brainstorm as a group. Sprint method explicitly avoids this; if the team will not commit to independent work, the sprint will not produce diverse sketches and the heat-map becomes a popularity contest.
  • The sketch step is being treated as the prototype. Sketches are concept exploration on paper or Figma frames, not Thursday's working prototype.
  • Sketches are due Tuesday morning. Tuesday is the full day; rushing to lunch produces sketches the team cannot read Wednesday.

What This Skill Produces

A single bundled artifact with four sections:

  1. Lightning demo board: each team member presents 3 demos (3 minutes each); the Facilitator extracts the reusable pattern from each demo into a one-line note. Result is a board with 12-21 patterns (4-7 team members x 3 demos).
  2. Sketch assignment plan: divide (each person sketches a different part of the target moment) or swarm (everyone sketches the same target). Default is swarm for v0.1 Design Sprints; divide is for teams running their second sprint with the same target who want to broaden coverage.
  3. Four-step sketches from each team member: Notes (20 min reviewing Monday output + lightning demos), Ideas (20 min rough doodles), Crazy 8s (8 minutes, 8 variations of the strongest Idea), Solution Sketch (30-90 min final 3-panel storyboard-style sketch).
  4. Recruiting tracker update: Riley (or the recruiter) confirms Friday slots; surfaces any cancellations and triggers buffer-slot activation if needed.

See references/TEMPLATE.md for the canonical structure and references/EXAMPLE.md for the Brainshelf book-catalog Tuesday artifact.

Tuesday Time Structure

The full Tuesday workshop is approximately 7 hours (09:00-12:30 + 13:30-17:00). The skill's bundled artifact emerges across the day:

  • 09:00-09:15: Welcome + Monday recap + Tuesday agenda
  • 09:15-10:45: Lightning demos (each person 3 demos x 3 min + 1 min Facilitator extraction = ~12 min per person; 4-person team = ~50 min; 7-person team = ~85 min)
  • 10:45-11:00: Break
  • 11:00-11:30: Sketch assignment plan + Notes step (each person silently reviews Monday output and lightning demo board)
  • 11:30-12:30: Ideas step (each person silently doodles ideas in their notebook or Figma frame)
  • 12:30-13:30: Lunch + (in parallel) recruiting tracker check-in
  • 13:30-13:40: Crazy 8s (8 minutes, 8 variations of strongest Idea)
  • 13:40-15:30: Solution Sketch (final 3-panel sketch; everyone silent; Facilitator times)
  • 15:30-15:45: Break
  • 15:45-16:30: Sketches collected + named + photographed for Wednesday board; sketcher attribution hidden (Wednesday's heat-map is blind)
  • 16:30-17:00: Day-end check-in; confirm sketches uploaded; confirm Wednesday morning attendance

This skill's 180-minute timebox covers the Facilitator-led portions (lightning demos + assignment + collection). The silent sketch steps (Notes, Ideas, Crazy 8s, Solution Sketch) are individual work and run in parallel; the team is together in the room (or video call) but not collaborating.

Common Pitfalls

  • Group brainstorming. The single most common failure mode. The Sprint method explicitly forbids group sketching because it produces sketches that converge on the loudest voice, not the best idea. Facilitator's job is to enforce silence during sketch steps.
  • Sketches not concrete enough. A solution sketch must be understandable Wednesday morning WITHOUT the artist explaining it. If the heat-map team has to ask "what is this?", the sketch failed the readability bar.
  • Skipping lightning demos. Lightning demos are not optional warm-up; they inject outside patterns the team would not have generated. Skipping them produces inward-looking sketches.
  • Demo facilitator failing to extract reusable patterns. Each demo's reusable pattern must be captured (not the whole demo). "Spotify's home screen" is not a pattern; "feed of personalized cards with one-tap action" is.
  • Sketching the prototype. The solution sketch is a concept; the prototype is Thursday's build. Sketches that try to be the prototype are over-detailed in the wrong dimensions (visual polish, copy) and under-detailed in the right ones (interaction sequence, user reaction).
  • Sketcher attribution leaked into Wednesday heat-map. Wednesday's heat-map is blind (no sketcher names visible) so the team votes on the sketch, not the sketcher. If sketches arrive on Wednesday with names attached, the Facilitator strips them.

Cross-Skill Usage

Prerequisites: tool-design-sprint-map-and-target. Sketch consumes Monday's target moment as the design target for the four-step sketch protocol. Without a target moment, sketches diverge with no shared direction.

This skill does NOT invoke tool-note-and-vote. Tuesday has no voting moment; all voting happens Wednesday after sketches are produced.

Next invocation in the sprint: tool-design-sprint-decide-and-storyboard Wednesday morning.

Canonical Sources

Decider Checkpoint

This skill ends with a Decider Checkpoint in references/TEMPLATE.md. There is NO mid-day Tuesday decision; the Decider's role Tuesday is to be a sketcher among other sketchers (the Decider sketches independently like everyone else). The end-of-Tuesday Decider Checkpoint is a logistics confirmation: all sketches collected, attribution stripped, Wednesday morning attendance confirmed. The substantive Decider call comes Wednesday morning at the supervote.

GitHub 仓库

product-on-purpose/pm-skills
路径: skills/tool-design-sprint-sketch
0
agent-skillsai-skillsclaude-codeclaude-desktopdesign-sprintfoundation-sprint

相关推荐技能

content-collections

Content Collections 是一个 TypeScript 优先的构建工具,可将本地 Markdown/MDX 文件转换为类型安全的数据集合。它专为构建博客、文档站和内容密集型 Vite+React 应用而设计,提供基于 Zod 的自动模式验证。该工具涵盖从 Vite 插件配置、MDX 编译到生产环境部署的完整工作流。

查看技能

polymarket

这个Claude Skill为开发者提供完整的Polymarket预测市场开发支持,涵盖API调用、交易执行和市场数据分析。关键特性包括实时WebSocket数据流,可监控实时交易、订单和市场动态。开发者可用它构建预测市场应用、实施交易策略并集成实时市场预测功能。

查看技能

creating-opencode-plugins

该Skill帮助开发者创建OpenCode插件,用于接入命令、文件、LSP等25+种事件。它提供了插件结构、事件API规范和JavaScript/TypeScript实现模式,适合需要拦截操作、扩展功能或自定义事件处理的场景。开发者可通过它快速构建响应式模块来增强OpenCode AI助手的能力。

查看技能

sglang

SGLang是一个专为LLM设计的高性能推理框架,特别适用于需要结构化输出的场景。它通过RadixAttention前缀缓存技术,在处理JSON、正则表达式、工具调用等具有重复前缀的复杂工作流时,能实现极速生成。如果你正在构建智能体或多轮对话系统,并追求远超vLLM的推理性能,SGLang是理想选择。

查看技能