apple-health-plugins

apple-health-analysis

A Claude Code plugin for evidence-based Apple Health data analysis. Transforms raw Apple Health export.xml into structured CSV data, then applies tier-based analysis with research-backed risk positioning.

Features

• Three-phase ETL pipeline: XML → CSV → Derived metrics → Evidence-based analysis
• Tier-based indicator system: Indicators classified by measurement reliability and clinical evidence strength
• Research-backed risk positioning: Dose-response mapping with cited epidemiological studies
• Parallel analysis agents: 5 specialized subagents for concurrent analysis
• Zero speculation: All interpretations anchored to specific research citations

Installation

As a plugin (from marketplace)

/plugin marketplace add github:loki/apple-health-analysis
/plugin install apple-health-analysis@apple-health-plugins

As a plugin (direct)

/plugin install --plugin-dir /path/to/apple-health-analysis

Usage

1. Export your Apple Health data from the Health app on iPhone (Settings → Health → Export All Health Data)
2. Place export.xml in an accessible path
3. Use the /apple-health-analysis skill:

/apple-health-analysis

Or ask Claude directly:

Analyze my Apple Health export at /path/to/export.xml

Pipeline

Phase	Description	Output
0: ETL	Single-pass XML → CSV extraction	~35 CSV files
1: Derived	Aggregated metrics computation	11 derived datasets
2: Analysis	Parallel tier-based analysis	Per-indicator reports
3: Synthesis	Cross-indicator report	Final health report

Indicator Tiers

Tier	Indicators	Analysis Depth
Tier 1	Steps, Resting HR, VO2max, Sleep, Weight/BMI	Full A/B/C/D
Tier 2	HRV, SpO2, ECG, Walking Steadiness	A/B/C + qualitative D
Tier 3	Energy, Respiratory Rate, Gait, Workouts, Noise	B/C only

Plugin Structure

apple-health-analysis/
├── .claude-plugin/
│   ├── plugin.json
│   └── marketplace.json
├── skills/
│   └── health-analysis/
│       ├── SKILL.md
│       ├── scripts/
│       │   ├── parse_health_v2.py
│       │   └── build_derived_v2.py
│       └── references/
│           ├── research-citations.md
│           ├── measurement-accuracy.md
│           └── analysis-framework.md
├── agents/
│   ├── tier-1-analysis.md
│   ├── tier-2-analysis.md
│   ├── tier-3-analysis.md
│   ├── data-quality.md
│   └── correlations.md
├── LICENSE
└── CHANGELOG.md

Requirements

• Python 3.8+ (for ETL scripts)
• numpy and pandas (for derived metrics)
• Claude Code with plugin support

License

MIT