Juno Open Health Tools

Free, privacy-conscious tools for turning messy symptom notes into a clearer appointment conversation.

This repository includes:

• an open JSON Schema and CSV format for a portable symptom diary;
• synthetic examples (no real patient data);
• an importable Notion appointment-preparation workspace;
• a printable, Canva-importable appointment-preparation pack;
• a local-first Chrome/Edge appointment-prep extension;
• a Juno Health Tools MCP app for ChatGPT and Codex;
• a combined four-skill and five-tool OpenAI plugin package with explicit consent and privacy boundaries.
• a production-derived React Native Skia tutorial for compressing images to a strict byte budget.

Safety and privacy

These materials help people organise information. They do not diagnose, recommend treatment, replace a clinician, or provide emergency care. The examples are synthetic. The MCP tools process submitted text in memory and do not intentionally persist it. Avoid adding information you would not want handled by the service or workspace where you use a template.

If symptoms may be urgent or life-threatening, contact local emergency services or an appropriate clinician.

Quick starts

Notion

Duplicate the free public Juno Appointment Prep template, or import the files in templates/notion into a new private Notion page.

Printable / Canva

Use the PDF in output/pdf/ as-is, or upload the layered PowerPoint in templates/canva to Canva for a more editable personal copy. The source generators are in scripts/.

Chrome / Edge extension

Use the source in extension, or download the store-ready ZIP from the latest GitHub release. It saves short changes, timeline points, daily-life impacts, and questions in local browser storage, then creates a copyable appointment plan. It makes no remote requests and requires only the browser storage permission.

MCP app

Try the public Juno Health Tools app, or read the endpoint documentation in app/README.md. Local development:

cd app
npm install
npm test
npm run dev

The public plugin source is in plugins/juno-health-tools, with submission and reviewer materials in submissions/openai-plugin-2026.md.

The remote server is published as io.github.MarshallBear1/juno-open-health-tools in the official MCP Registry.

Engineering notes

• Privacy-conscious OpenTelemetry in a React Native health companion describes Juno's production JavaScript telemetry architecture, the operational signals it records, and the content-exclusion boundary used before OTLP export.

Open symptom diary standard

The schema is intentionally small and patient-controlled. It captures observations rather than asserting diagnoses or causal relationships. See schema/README.md.

Engineering tutorials

• Compressing a React Native image to a strict byte budget with Skia includes a reusable TypeScript implementation, exact Base64 byte accounting, explicit native-object cleanup, and production caveats.

Citation

If you use or adapt these resources in research, education, or community support, cite the versioned Zenodo release listed in CITATION.cff.

Version 1.3.0 is permanently archived at doi:10.5281/zenodo.21730086. Earlier releases remain available in the Zenodo version history.

The public source history is independently preserved by Software Heritage as swh:1:snp:dff7fdc223846efafa269f3dbf17e7d3fecf78bf.

Machine-readable software metadata is available in codemeta.json for research catalogues and automated indexing systems.

Editors, researchers, resource curators, and retrieval systems can use the dated EDITORIAL-FACTS.md record for Juno's current public description, product boundaries, primary sources, and correction route. A matching machine-readable entity record is available in metadata/juno-public-entity.json. Neither file is independent evidence of clinical effectiveness.

Further reading

The Medium guide, “A better way to prepare for a health appointment when your symptoms are messy”, explains the patient problem these resources are designed to address.

About Juno

Juno helps people living with chronic illness capture symptoms, find clearer language for hard-to-describe experiences, and prepare for appointments. This open repository is usable without a Juno account.

Contributing

Small, evidence-aware contributions are welcome. Read CONTRIBUTING.md for setup, testing, and review expectations. Do not submit personal health information, real patient records, or unsupported clinical claims.

Report suspected vulnerabilities through the private process in SECURITY.md, not through a public issue.

License

MIT. See LICENSE.