The first version keeps you in control: it drafts the visit record, home exercises, and this week’s plan, then waits for review before writing anything into your tracker.
- Structured PT encounter saved with raw notes plus cleaned summary
- Suggested weekly PT program linked into your existing exercise library
- Per-exercise “how to perform it” coaching text and video search links
- Data shape that can later be shared into Mike’s health app
Paste text from your PT (handwritten notes, email, etc.). The parser extracts exercise names plus sets, reps, and time into draft entries you can review before saving.
Select a PDF, plain-text, or .json file. PDFs are text-extracted automatically. JSON files matching the template below are parsed directly.
Paste this template into ChatGPT, Claude, etc. Ask it to fill in the exercises your PT prescribed. Then paste the returned JSON into the text box above and click Parse Exercises.
{
"category": "",
"exercises": [
{
"exerciseName": "",
"targetBodyPart": "",
"prescribedFor": "",
"type": "",
"description": "",
"sets": "",
"reps": "",
"timeSec": "",
"youtubeVideoUrl": "",
"notesAndCues": ""
}
]
}
Describe your symptoms and I will suggest matching exercises from your library. Example: "lower back pain with sciatic nerve tightness when sitting."
This tab is the first plumbing pass for the shared health AI layer. It builds a sanitized payload, estimates cost, and sends the request to Supabase Edge Functions when configured.
Format: exercise, date, sets, reps, weight, side