Getting Started
SpeziVibe is a collection of installable skills for AI coding tools. Each skill walks you through a specific digital health planning area — needs analysis, compliance, data modeling, UX, study design — and produces a structured markdown document you can build from.
Installation
Install all skills with a single command:
npx skills add StanfordSpezi/SpeziVibe --all
Skills are tool-agnostic. They work with Claude Code, Cursor, GitHub Copilot, OpenAI Codex, Gemini CLI, and any other tool that supports installable skills or custom instructions.
Quick Start
Once installed, start by describing what you want to build. The build-an-app skill figures out which planning steps apply and walks you through each one interactively:
- Describe your app — Tell your AI coding tool what you want to build
- Work through each skill — Each relevant skill asks you questions and produces a planning document (e.g.,
need-statement.md,compliance-brief.md,ux-brief.md) - Get an implementation plan — The app-build-planner reads your planning docs and produces a milestone-based build plan
What's Included
SpeziVibe includes skills organized into planning and release categories:
Planning Skills
- build-an-app — Orchestrates the other skills based on what you describe; walks you through each one in order
- biodesign-needs-finding — Guided needs-finding using the Stanford Biodesign process; produces
need-statement.md - spezi-platform-selection — Helps choose between React Native and Apple-native, sets up your dev environment, and clones the starter repo
- digital-health-study-planning — Plans a research protocol including enrollment, consent, assessments, and outcomes; produces
study-brief.md - digital-health-compliance-planning — Walks through HIPAA, IRB, FDA, GDPR applicability and recommends controls; produces
compliance-brief.md(not legal advice) - health-data-model-planning — Defines health data entities, relationships, and FHIR fit; produces
data-model-brief.md - digital-health-ux-planning — Plans user journeys, onboarding, and engagement for patients and clinicians; produces
ux-brief.md - fhir-data-model-design — Maps clinical data to FHIR R4 resources with terminology bindings and sample JSON; produces
fhir-data-model.md - app-build-planner — Reads planning docs and produces a milestone-based implementation plan; produces
implementation-plan.md
Release Skills
- keep-a-changelog-generator — Generates changelog entries from git history in Keep a Changelog format
- release-notes-generator — Creates user-facing release notes with features, fixes, and migration guidance
Next Steps
- Browse the Skills reference for detailed descriptions of each skill
- Visit the GitHub repository for source code and contributions