How to Export Your Fitness Data as a Spreadsheet or Full Backup

To export your fitness data from ChatRPE, ask for it in the chat. "Export my data" returns a full backup: a zip with your entries as a CSV, your notes as Markdown, and an index of everything you have tracked. To get only part of your history, say which part and which file: "Export my weigh-ins from this year as a spreadsheet" returns one CSV of those entries, with a column per field.

Exporting copies your history. It does not move, change, or delete anything.

Name the slice and the file

The coach turns your request into filters, so describe the export in your own words. It can narrow by:

• Category, such as workouts, meals, weigh-ins, sleep, or any type you invented. Several can go in one file.
• Dates, as a range of days in your local time ("March 1 through May 31", both days included) or a window such as the past week, month, or year.
• A word that appears in an entry's fields, note, or category: "everything that mentions my knee".
• Count and order: "my latest 50 workouts", or "my first 100 entries, oldest first". Newest first is the default.

Then choose the file:

Only the zip carries your profile, notes file, dashboard list, and the context the coach keeps about you. A CSV or JSON export holds entries alone.

If a filter matches nothing, no file is made and your export allowance is untouched. The coach sees which filters found nothing and can suggest a wider request.

Where to ask, and how often

The file is assembled in your browser from your own signed-in session, so ask in a chat on the web or in the iPhone app. Siri, or a slow reply the app hands off to the server, comes back with a refusal asking you to retry from the app.

In a browser the download starts on its own. In the app, a card with a Save button appears in the reply and opens the share sheet. The file is released after about five minutes, so save it while you are still in the conversation.

Free and Basic accounts can export once a week and Premium once a day. A filtered export counts the same as a full one, so a weigh-in CSV on Monday uses the week's export on a Free plan. When a Save tap is required, the window starts when you save, and dismissing the share sheet does not spend it.

What the full backup holds

notes.md reads like a document rather than a database dump, which is what a coach or training partner wants. Its list of annotated entries stops at 500 and points to the CSV for the remainder. The workout journal guide covers writing notes worth reading later.

Open index.json first if you plan to parse anything. It names every data type in the file, how many entries each holds, the date of the most recent one, and each field path with the value types seen there. ChatRPE has no fixed schema, so these paths describe whatever you and the coach actually logged.

Reading workouts in a spreadsheet

By default the CSV has five columns: id, data_type, date (an ISO 8601 timestamp), notes, and data_json, which holds the entry's fields as JSON in one cell. A weigh-in's payload is little more than a number and a unit. A lifting session holds a tree:

{"exercises":[{"name":"Bench Press","sets":[{"weight":205,"reps":6,"rpe":8},{"weight":205,"reps":6,"rpe":8.5}]}]}

Asking for one column per field replaces data_json with a column for each field, so weigh-ins, sleep hours and custom metrics open as ordinary columns you can sort and chart. A field nested inside another gets a dotted name, such as a.b. Lists stay whole: the exercises above lands in one cell as JSON, because spreading it out would give each session a different set of columns.

So a set-by-set table still takes a script, Power Query, or a JSON tool. index.json lists the paths to pull, such as exercises[*].sets[*].rpe, where [*] marks a level that repeats. If another program will read the data, the JSON format gives it every entry in full without CSV quoting in the way.

Expect one quirk. Cells whose text begins with =, +, -, @, or a tab are written with a leading apostrophe so a spreadsheet shows them as text instead of running them as a formula. A note reading "-2.5 lb since Monday" appears as '-2.5 lb since Monday. Plain numbers are exempt and stay numeric. Strip the apostrophe if you parse the file yourself.

Download a chart when the export is spent

Every chart ChatRPE draws in a reply has a Download CSV button in its top corner, and chart widgets on a dashboard offer the same control. Chart downloads are not rate limited, so they are the way to pull one number out midweek after your export is used.

"Chart my bodyweight for the last six months."

The file holds the plotted series, not your entries. Points are combined to one per day, so set-level detail is not there. A chart reads up to 1,000 entries of a type in the period, and a long series is thinned to about 200 points that keep its shape, so a year of daily weigh-ins comes out with gaps. The date column uses the chart's labels rather than ISO timestamps. For every raw entry in a range, use a filtered export instead.

Taking a record to an appointment

Decide the question before handing anyone a file. A full backup carries far more than a clinician or dietitian needs. A short summary with the relevant entries is usually more useful:

"Summarize my calories, protein, bodyweight trend, and workout attendance for the last three months, and flag gaps in the record. Then export my meals and weigh-ins from those three months as a spreadsheet with one column per field."

That gives you the reading and one file of the underlying entries. Medical targets and treatment decisions belong with your clinician. What the file adds is a record to work from, including its gaps, rather than a recollection assembled in the waiting room.