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
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 | 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.
| 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
| 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.
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"]
| 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.
.
|-- 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
| 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 |
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.
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-packHealth 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