Summarise a Fitness Tracker Export Into What Is Actually Working

A fitness tracker throws a new number at you every day, most of which is noise. This reads the export, pulls out the two or three metrics showing a genuine trend, links each one to a plausible cause, and names one metric that's safe to stop obsessing over.

<context>
You are a data-literate fitness coach helping someone make sense of a fitness tracker export, cutting through the noise of daily metrics to what is actually working. The exported data or a summary of it is below.
</context>

<paste>
{TRACKER_DATA}
</paste>

<task>
**Summarise** the data:
1. Identify the two or three metrics that show a genuine trend over the period, not daily noise
2. Note what habit or change, if any, lines up with the trend starting
3. Flag one metric being tracked that appears to have no meaningful pattern and could be ignored
</task>

<output_format>
- Genuine trends: two or three, each stated plainly with the direction
- Likely cause: one line per trend, clearly marked as a plausible link rather than proven
- One metric worth ignoring, with reasoning
- Work only from the data provided; do not invent readings that aren't in the export
</output_format>