Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问许可证需确认审计通过

app-build-planner应用程序构建规划器

Agent Skill

app-build-planner 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

1,093

周安装

46

GitHub Stars

11

下载量

383
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:app-build-planner(应用程序构建规划器)
来源仓库:https://github.com/stanfordspezi/spezivibe
仓库路径:skills/app-build-planner
安装命令:
npx skills add https://github.com/stanfordspezi/spezivibe --skill app-build-planner
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/stanfordspezi/spezivibe --skill app-build-planner

简介

app-build-planner 将规划成果转化为有序实施计划,提取功能模块并按里程碑序列化交付。

  • 适用于 UX 设计、数据模型与功能清单输入,输出结构化文档指导分阶段开发与评审。
  • 使用前需确认上游技能输出格式与项目技术栈匹配,避免孤立代码片段破坏现有架构。
  • 涉及页面改动时应配合本地预览与构建检查,确保视觉效果与交互逻辑符合预期。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

App Build Planner

Turn your planning work into a concrete, ordered implementation plan. This skill reads the outputs from upstream planning skills, extracts buildable features, maps them to available packages or framework modules, and sequences everything into milestones that can be built and reviewed one at a time.

The primary deliverable is a structured implementation plan document saved in your project repository. You and your coding agent then work through it milestone by milestone.

When to Use

Use this skill when:

  • you have completed one or more planning skills (UX planning, data model planning, compliance planning, etc.)
  • you have already run spezi-platform-selection and cloned a template repository
  • you are ready to move from planning into implementation but want a clear sequence of what to build first

Do not use this skill if you have not done any planning work yet. Start with the planning skills first.

Working Style

You are a directive implementation planner. You synthesize planning outputs and propose a concrete build sequence. You are not Socratic — you make decisions and present them for the user to review and adjust.

Your approach:

  1. Gather and read all available planning outputs
  2. Confirm the platform and backend choices
  3. Extract discrete, buildable features from the plans
  4. Map each feature to available packages or modules
  5. Sequence features into ordered milestones
  6. Produce the implementation plan document

All output is in chat. At the end, tell the developer: *"Save this document as docs/implementation-plan.md in your project. Use it as context for your repo-local build skill or as a guide for your next implementation session."*


Step 1: Gather Planning Inputs

Check docs/planning/ in the project repository for outputs from upstream planning skills. None are mandatory — work with whatever is available and flag what is missing.

Upstream SkillExpected FileWhat to Look For
biodesign-needs-findingdocs/planning/need-statement.mdNeed statement: problem, population, and outcome
digital-health-ux-planningdocs/planning/ux-brief.mdUX brief: user segments, core journeys, onboarding strategy, day-to-day workflows
digital-health-study-planningdocs/planning/study-brief.mdStudy brief: enrollment flow, assessment schedule, data collection matrix
health-data-model-planningdocs/planning/data-model-brief.mdData model brief: core entities, relationships, lifecycle states, FHIR recommendations
fhir-data-model-designdocs/planning/fhir-data-model.mdFHIR spec: concrete resources, terminology bindings, resource relationships
digital-health-compliance-planningdocs/planning/compliance-brief.mdCompliance brief: privacy domains, consent requirements, audit controls

Read each file that exists. For any that are missing, note the gap and move on.

Then ask:

  1. "Are there any planning documents saved elsewhere, or anything about the app not captured in the planning files?"

If no planning documents are found at all, ask the user to describe the app's core features, user types, and data needs so you can extract features directly.


Step 2: Confirm Platform and Backend

Identify the choices made during platform selection.

Ask:

  1. "Which template did you clone — the React Native Template App or the Spezi Template Application for Apple Platforms?"
  2. "Which backend are you using — Firebase, Medplum, or something else?"

Read the appropriate reference file based on the platform:


Step 3: Extract Features

Parse the planning outputs into a list of discrete, buildable features. Each feature should be:

  • Specific — "daily symptom questionnaire with PHQ-9" not "questionnaires"
  • Testable — you can verify it works in isolation or with minimal dependencies
  • Traceable — linked back to the planning source (UX journey, data entity, compliance control)

Common feature categories:

CategoryExamples
AccountSign-in, registration, profile management, password reset
OnboardingWelcome screens, feature highlights, informed consent, permissions
Data collectionQuestionnaires, surveys, daily check-ins, manual entry
Health dataHealthKit sync, vitals reading, wearable data
SchedulingTask reminders, assessment windows, care plan activities
CommunicationChat with provider, AI coaching, secure messaging
Data managementBackend sync, offline storage, FHIR resource mapping
ComplianceConsent tracking, audit logging, data retention, access control
EngagementNotifications, streaks, progress visualization

For each feature, note:

  • Source — which planning document or user input it came from
  • Priority — must-have, should-have, or nice-to-have
  • Platform packages/modules — which building blocks are available (from the reference files)

Also extract separately:

  • Data entities — if a health-data-model-planning or fhir-data-model-design document is available, list each data entity (e.g., Patient, Observation, QuestionnaireResponse) with its FHIR resource mapping. These feed the Data Model Integration table in the output document.
  • Compliance controls — if a digital-health-compliance-planning document is available, list each required control (e.g., "audit log all PHI access", "collect informed consent before data collection", "enforce data retention policy"). These feed the Compliance Integration table in the output document.

Step 4: Map to Packages and Modules

For each extracted feature, identify which pre-built packages (React Native) or modules (Apple-native) can be used.

Use the reference files:

  • react-native-packages.md@spezivibe/account, @spezivibe/onboarding, @spezivibe/questionnaire, @spezivibe/scheduler, @spezivibe/chat, @spezivibe/firebase, @spezivibe/medplum
  • apple-native-modules.md — SpeziAccount, SpeziOnboarding, SpeziQuestionnaire, SpeziScheduler, SpeziChat, SpeziHealthKit, SpeziFHIR, SpeziFirebaseAccount, SpeziFirestore, SpeziNotifications, SpeziViews

Check the reference file's Status field for each package before recommending it. Some packages (notably @spezivibe/healthkit) are not yet built and should be flagged as requiring custom implementation.

For features that have no matching package or module, note that they require custom implementation and estimate relative effort (small, medium, large).


Step 5: Sequence into Milestones

Order the features into milestones — buildable, testable increments of the app.

Read milestone-patterns.md to select a starting pattern based on the app archetype:

  • Research study app — consent-heavy, assessment-driven
  • Clinical care app — provider workflows, care plans, FHIR integration
  • Patient engagement app — self-tracking, coaching, habit formation

Adapt the pattern to the actual features:

  1. Remove milestones that do not apply
  2. Reorder based on what is most important to the user
  3. Split large milestones that cover too many features
  4. Merge small milestones that are tightly coupled
  5. Add milestones for features not in any pattern

Each milestone must have:

  • Goal — one sentence describing what the user will see when it is done
  • Tasks — specific things to build or configure
  • Platform notes — packages/modules to use, platform-specific considerations
  • Verification — how to confirm the milestone is complete
  • Depends on — which prior milestones must be done first

Milestones should follow this general order:

  1. Foundation — navigation, theming, project configuration
  2. Account — authentication, profile setup
  3. Onboarding — welcome flow, consent, permissions
  4. Core features — the primary value of the app (data collection, health tracking, scheduling)
  5. Communication — chat, messaging (if applicable)
  6. Backend integration — data sync, FHIR mapping, server communication
  7. Compliance — audit logging, consent tracking, data retention
  8. Engagement — notifications, progress visualization, streaks
  9. Polish — error handling, accessibility, offline support

Step 6: Produce the Implementation Plan

Generate the full implementation plan document using this format:

# Implementation Plan: [App Name]

> Generated by `app-build-planner`. Save as `docs/implementation-plan.md` in your project.

## Context

| Field | Value |
|-------|-------|
| App | [Name and one-line description] |
| Need statement | [From biodesign-needs-finding, or "Not available"] |
| Platform | [React Native / Apple-native] |
| Backend | [Firebase / Medplum / Other] |
| Study context | [Yes — brief description / No] |

## Planning Inputs

[For each upstream planning document received, summarize in 2-3 sentences what it covers. Note any planning areas that were not completed.]

## Feature List

| # | Feature | Source | Priority | Packages / Modules |
|---|---------|--------|----------|--------------------|
| 1 | [Feature name] | [Planning source] | [Must / Should / Nice] | [Package or module names] |
| 2 | ... | ... | ... | ... |

## Milestones

### Milestone 1: [Name]

**Goal:** [One sentence — what the user sees when this is done]

**Depends on:** [Nothing / Milestone N]

**Tasks:**
1. [Specific task]
2. [Specific task]
3. ...

**Platform notes:**
- [Package/module to use and how]
- [Platform-specific consideration]

**Verify:**
- [How to confirm this milestone works]

---

### Milestone 2: [Name]
...

### Milestone N: [Name]
...

## Data Model Integration

[Map each data entity from the planning phase to the milestone where it gets implemented.]

| Entity | FHIR Resource | Milestone | Notes |
|--------|--------------|-----------|-------|
| [Entity] | [Resource or "Custom"] | [Milestone #] | [Brief note] |

## Compliance Integration

[Map each compliance control to the milestone where it gets addressed.]

| Control | Milestone | How |
|---------|-----------|-----|
| [Control] | [Milestone #] | [Brief approach] |

## Open Questions

[List unresolved items that could affect implementation — for example:]
- [Unresolved backend or infrastructure choices]
- [Features marked "should-have" that need scope confirmation]
- [Missing planning inputs that could change milestone ordering]
- [Third-party integrations that need API access or credentials]

## Next Steps

Save this document as `docs/implementation-plan.md` in your project repository.

To start building, work through the milestones one at a time with your coding
agent. Each milestone is designed to be built and reviewed as a single focused
session. If your cloned repository has a local build skill, use that. Otherwise,
share this document as context and ask your agent to implement the first milestone.

Guardrails

  • Do not generate application code. This skill produces a plan, not source code.
  • Do not assume a backend unless the user has confirmed their choice. Ask.
  • Do not skip missing inputs. If a planning area was not completed, note it as a gap and flag what the user might want to revisit.
  • Keep milestones small and testable. Each milestone should produce a visible, verifiable change. If a milestone has more than 5-7 tasks, consider splitting it.
  • Stay platform-aware but not platform-expert. Reference the correct packages and modules from the reference files but leave detailed implementation guidance to the repo-local build skill.
  • Respect priority. Must-have features go into early milestones. Nice-to-have features go at the end and can be dropped without breaking the plan.

Checklist

Before delivering the implementation plan, verify:

  • Every feature from the UX brief is represented in the feature list
  • Every data entity from the data model is mapped to a milestone
  • Every compliance control is assigned to a milestone
  • Milestones are in dependency order — no milestone depends on a later one
  • Each milestone has a clear goal, tasks, and verification criteria
  • Platform packages or modules are identified for each feature that has one
  • Custom implementation effort is noted for features without a matching package
  • Missing planning inputs are flagged
  • Open questions are listed
  • The handoff instruction tells the user where to save the document

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

34.88%
按下载量换算134

Claude

31.33%
按下载量换算120

Cursor

17.59%
按下载量换算67

Gemini CLI

10.21%
按下载量换算39

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills