draft-flow
安装方式
先用 Prompt 核对来源,也可切换到安装命令或下载可用的安装包。
用 AI 助手安装
复制提示词,让助手先核对固定版本和依赖。
查看 Prompt收起 Prompt
先检查当前 Agent 是否已安装 SkillPI Skill Finder。如果尚未安装,请按官方仓库的说明安装并启用:https://github.com/Findskills/skillpi。然后使用 Skill Finder 打开此 SkillPI 详情页并定位 "draft-flow":https://skillpi.com/zh/skills/deepskill-0db13c94d9f74346cfd550b02ba1da8e1bad351bfdf44deba3df2e458c02f76f。安装前核对上游来源和版本:https://github.com/jeremylongshore/tons-of-skills-marketplace/blob/6ea46b08dce892944db7c16f28d0b5308d402402/plugins/ai-agency/tonone/skills/draft-flow/SKILL.md。来源不一致时停止并说明差异。核对通过后,用 Skill Finder 安装目标 Skill,并报告安装路径和来源版本。
此命令从 GitHub 来源仓库获取指定技能。运行前请先查看源文件。
npx skills add https://github.com/jeremylongshore/tons-of-skills-marketplace/tree/6ea46b08dce892944db7c16f28d0b5308d402402/plugins/ai-agency/tonone/skills/draft-flow --skill draft-flow来源与配置
查看 GitHub 源码jeremylongshore/tons-of-skills-marketplace固定提交 6ea46b08dce892944db7c16f28d0b5308d402402
配置方式待核实
请查看原始来源中的适用工具、依赖和权限要求。
安装实测待补充
本条目收录公开技能“draft-flow”,按来源分类归入「做封面图与设计」使用场景、「设计、媒体与内容创作」职业。目录标签:ai-lab、用户流程、线框图、流程图。此页用于按主题发现相关资源;请通过固定原站版本核对实际能力、依赖环境和所需权限。本站仅提供独立目录介绍与原站入口,原始文件和附件依各自许可开放。
文件浏览
---
name: draft-flow
description: |
Use when asked to design a user flow, map how a user moves through a feature, create a wireframe or flow diagram, or document interaction design for a product brief. Examples: "design the flow for X", "map out the user journey", "create a wireframe for this feature", "how should the UX work for this".
allowed-tools: Read, Write, Edit, Bash, Glob, Grep, WebFetch, WebSearch, Task, TodoWrite, AskUserQuestion
version: 0.6.4
author: tonone-ai <hello@tonone.ai>
license: MIT
---
# Draft Flow
You are Draft — the UX designer on the Product Team.
Follow the output format defined in docs/output-kit.md — 40-line CLI max, box-drawing skeleton, unified severity indicators, compressed prose.
## Steps
### Step 1: Understand the Job
Read the input — a product brief from Helm, a feature description, or a user task. Identify:
- **Primary task:** What is the user trying to accomplish?
- **Starting state:** Where is the user when this task begins? (logged out? empty state? mid-session?)
- **Done state:** What does "task complete" look like from the user's perspective?
- **User's mental model:** What does the user already know/expect going in?
If working from a Helm brief, map `success_criteria` to the done state directly.
### Step 2: Map the Happy Path
Produce a Mermaid flowchart for the primary success path. Label nodes with the user's action or decision, not UI element names.
```mermaid
flowchart TD
A[User arrives at...] --> B{Decision point}
B -->|Option A| C[User does...]
B -->|Option B| D[User does...]
C --> E[Task complete]
```
Rules for the happy path:
- Every node is a user action or system response — no "page" nodes
- Every diamond is a decision the user must make — label both branches
- The start node states where the user is and what triggered the task
- The end node states what the user sees and knows at completion
### Step 3: Add Error and Empty States
Extend the diagram with:
- **Validation errors** — what happens when user input is wrong? Where do they land?
- **Empty states** — what does the user see on first use, before they have data?
- **Dead ends** — every error must have a recovery path; no flow should end without a resolution
Mark error/empty paths in the diagram with `:::error` or a note annotation.
### Step 4: Annotate Decision Points
For each diamond (decision fork) in the flow, add an annotation:
```
[Decision: "Do they have an account?"]
Context: User may arrive from a marketing link without a session.
What they need: Clear indication of whether sign-in or sign-up is the right path.
What we provide: [describe what the UI shows at this point]
Risk: [what goes wrong if we get this wrong]
```
### Step 5: Identify Friction Points
Review the full flow. Flag any step where:
- The user must recall information they weren't given earlier in the flow
- The user must make a decision without enough context
- A single error forces the user to restart from the beginning
- The flow requires more than 3 consecutive user actions without system feedback
Mark these with `▲ FRICTION:` annotations.
### Step 6: Deliver
Present:
1. The Mermaid flow diagram (full, renders cleanly)
2. Annotated decision points
3. Friction flags with recommended resolutions
4. One-paragraph summary of the key UX decisions made and why
## Delivery
If output exceeds the 40-line CLI budget, invoke `/atlas-report` with the full findings. The HTML report is the output. CLI is the receipt — box header, one-line verdict, top 3 findings, and the report path. Never dump analysis to CLI.
当前收录版本
0.6.4
质量评测
满分 100。前四项来自 DeepSkill 收录的质量报告。
评分依据
专注于 UX 设计流程与用户旅程映射,解决产品设计领域的明确痛点
评分依据
触发词覆盖全面,示例清晰,边界明确
评分依据
步骤划分清晰,包含端到端示例和交付标准
评分依据
流程设计合理,无明显执行漏洞
计算方式
取前四项原始分值的算术平均,四舍五入取整。可通过收录报告核对各项评分与说明,运行记录需单独查证。
权限与使用前检查
根据当前主文档自动整理,命中内容可能是示例或防护说明;不是实际权限声明,也未覆盖附件、外部依赖或运行行为。
未从主文档识别到上述常见操作,不能据此判断无需权限或没有风险。
安全评级与访问范围
risk · 静态证据核验,不代表运行安全保证。
使用前请由宿主核对并授权实际文件读写、外部服务和联网目的地;本站的核验不会授予任何运行权限。
用户讨论
还没有已核实的点评。