Skip to content

Navigation Menu

Sign in
Sign up

Latest commit

History

12 Commits

Folders and files

NameName
Last commit message
Last commit date

Repository files navigation

Health Intelligence System banner showing a private health data vault connected to wearable telemetry, clinician handoff workflows, and research-only intelligence layers.

Health Intelligence System

A public, safety-gated protocol and agent pack for private health operations, clinician-ready handoffs, wearable-data organization, cancer-prep workflows, and research-boundary discipline.

Release License Safety Verified Package Agent Pack

Download full package · Download agent pack · Quick start · Safety · Distribution

What This Is

Health Intelligence System is a public protocol layer for becoming a better health operator and patient advocate. It helps organize private records, wearable exports, doctor visits, clinician questions, evidence sources, cancer-prep workflows, and AI-assisted summaries without pretending to be a doctor.

It is designed to be used in three ways:

Mode Use it for Output
Human operator Private vault setup, records, visits, routines, questions Templates and checklists
Coding agent / assistant Repeatable workflows for setup, redaction, visit prep, handoff, weekly review Commands, prompts, plugin skill
Product runtime Public safety contract consumed by private Agentic Life OS / Agentic Health OS Protocol, package, validation gates

Vertical Packs

Vertical Audience Public core Private / licensed layer
Gut Intelligence System Parents, qualified nutrition teams, family-health products Family context, safety gates, ordinary meal journey, weekly review, coach handoff, evidence receipts Microbiome interpretation, proprietary science mappings, identifiable family state, branded product and coaching operations

The gut vertical includes the installable gut-family-journey skill, machine-readable schemas, fictional fixtures, and executable safety evaluations. It does not interpret raw microbiome data or provide medical nutrition therapy.

It does not diagnose, interpret labs or imaging, prescribe, choose treatment, change medication, dose supplements, triage emergencies, or replace clinicians.

Download

Artifact Best for Link Verify
Full release package Reviewers, builders, product integration, complete documentation health-intelligence-system-v0.2.1.zip release-manifest.json
Agent pack Codex, Claude, ChatGPT, OpenCode, local assistants, vault setup health-intelligence-agent-pack-v0.2.1.zip agent-pack-manifest.json
Source repo GitHub review, issues, pull requests, forks frankxai/health-intelligence-system Git history

Local verification:

npm run package:release
npm run verify:release
npm run package:agent-pack
npm run verify:agent-pack

Verify the public downloads:

pwsh -NoProfile -ExecutionPolicy Bypass -File scripts/verify-release.ps1 -Version 0.2.1 -Download
pwsh -NoProfile -ExecutionPolicy Bypass -File scripts/verify-agent-pack.ps1 -Version 0.2.1 -Download

Install Surfaces

Surface Path Install or adapt into
Codex plugin plugins/health-intelligence-system/ Codex plugin directory
Sovereign Health Operator skill plugins/health-intelligence-system/skills/sovereign-health-operator/ Codex, local skill runners
ChatGPT Project prompt prompts/chatgpt-project-system-prompt.md ChatGPT Projects
Custom GPT instructions prompts/custom-gpt-instructions.md Custom GPT builder
Claude Project prompt prompts/claude-project-prompt.md Claude Projects
Local redaction prompt prompts/local-llm-redaction-prompt.md Local LLM privacy review
Slash commands commands/ Coding agents, local automation
Private vault templates templates/ Obsidian, local folders, encrypted workspace

Start with AGENT_PACK.md if you want the installable operator package instead of the full repository.

System Architecture

flowchart LR
 user["Person / patient advocate"] --> vault["Private health vault"]
 vault --> records["Records, visits, routines, wearable exports"]
 records --> his["Health Intelligence System public protocol"]
 his --> agent["Agent pack: plugin, prompts, commands, templates"]
 agent --> handoff["Clinician handoff and question packets"]
 his --> safety["Safety, privacy, validation, review gates"]
 his --> alos["Private Agentic Life OS"]
 alos --> ahos["Agentic Health OS: personal health runtime"]
 alos --> lsr["Life Sciences Researcher IS: research-only"]
 lsr --> brief["Evidence brief and clinician questions"]
 brief --> handoff
 safety -. blocks .-> blocked["Diagnosis, test interpretation, treatment choice, medication changes, emergency triage"]
Loading

Product Boundary

Layer Public or private Job Must not hold
Health Intelligence System Public Protocol, release package, plugin, prompt pack, templates, safety contract Raw personal health records
Agentic Health OS Private product runtime Personal organization, nutrition/training/wellness loops, visit prep, private vault orchestration Public patient data
Life Sciences Researcher IS Private research-only package for now Literature, trials, mechanisms, biomedical evidence envelopes Personal records or care decisions
Clinician interface User-reviewed export Questions, timelines, source ledger, handoff packet Unreviewed diagnosis or treatment advice

Read the full boundary in docs/product-boundary.md and docs/agentic-life-os-integration.md.

Core Workflows

Workflow Command Template Boundary
Private vault setup commands/private-health-instance-setup.md templates/private-vault-manifest.md Public repo never stores real records
Doctor visit prep commands/doctor-visit-prep.md templates/doctor-visit-prep.md Questions, not diagnosis
Clinician handoff commands/clinician-handoff-export.md templates/clinician-handoff-export.md User-reviewed export only
Wearable ingestion commands/wearable-data-ingestion.md templates/wearable-data-ingestion-manifest.md Trends are not medical interpretation
Weekly review commands/health-optimization-weekly-review.md templates/health-operator-weekly-review.md Self-tracking, not prescription
Possibility map commands/health-possibility-map.md templates/health-possibility-map.md Possibilities to discuss, not answers
Privacy preflight commands/privacy-preflight-redaction.md templates/ai-sanitized-context-export.md AI redaction is never the only barrier
Cancer prep docs/cancer-detection-prep-treatment.md cancer templates in templates/ Care team owns clinical decisions

Repository Map

.
|-- assets/ # README banner and public visuals
|-- commands/ # Slash-command style operator workflows
|-- docs/ # Architecture, safety, evidence, integration, research boundary
|-- plugins/health-intelligence-system/
| `-- skills/sovereign-health-operator/
|-- prompts/ # ChatGPT, Custom GPT, Claude, local redaction prompts
|-- scripts/ # Release and agent-pack packaging/verification
|-- templates/ # Private-vault and clinician-handoff templates
|-- AGENT_PACK.md # Installable pack guide
|-- MARKETPLACE.md # Distribution and marketplace path
|-- SAFETY.md
|-- PRIVACY.md
|-- VALIDATION.md
`-- REVIEW-GATE.md

Documentation Index

Need Start here
Fast operator setup QUICK-START.md
Safety boundary SAFETY.md
Privacy model PRIVACY.md
Validation and release checks VALIDATION.md, VERIFY.md
Full architecture docs/architecture.md
Product split docs/product-boundary.md
Private runtime integration docs/agentic-life-os-integration.md
Research boundary docs/companion-research-systems.md
External systems comparison docs/external-systems-comparison.md
Agent installation docs/coding-agent-installation-guide.md
Prompt pack docs/prompt-pack-chatgpt-claude.md
Wearable data docs/wearable-data-ingestion-and-privacy.md
Cancer prep docs/cancer-detection-prep-treatment.md

Safety Contract

Not medical advice. This repository is an organizational and agent-workflow system for patient advocacy, privacy, records, questions, and clinician handoff.

Allowed:

  • organize records, timelines, questions, and routines;
  • summarize user-provided facts without adding medical conclusions;
  • prepare doctor-visit agendas and clinician handoff packets;
  • track source dates and evidence provenance;
  • create privacy-reviewed context exports for optional AI use;
  • translate research into questions for qualified clinicians.

Blocked:

  • diagnosis or reassurance that something is not serious;
  • lab, imaging, pathology, genomic, or wearable medical interpretation;
  • medication, supplement, dosing, fasting, rehab, or treatment recommendations;
  • emergency triage or advice to delay care;
  • storing private health records in public git history.

If a person has symptoms, abnormal tests, suspected cancer, confirmed cancer, severe side effects, or urgent concerns, this system routes them to qualified care.

Release Quality

Every public release should have:

  • versioned ZIP files;
  • checksums and manifests;
  • safety-critical docs in the package;
  • no secrets, private health data, or local-machine paths;
  • install path, first workflow, and support/upgrade path;
  • clinical/legal review gate clearly marked.

This repo uses:

npm run package:all
npm run verify:release
npm run verify:agent-pack

Built On

Health Intelligence System adopts the SIP file contract, sovereignty clause, attestation pattern, and composition discipline from Starlight Intelligence System.

The banner image was generated with the built-in GPT image generation workflow and saved into this repository at assets/health-intelligence-system-banner.png.

Built on SIP - Starlight Intelligence Protocol v1.1.1

About

SIP-aligned health intelligence system for cancer prep, treatment decision support, and survivorship records.

Topics

Resources

Contributing

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages

AltStyle によって変換されたページ (->オリジナル) /