See how your cardio has been trending

Three months of zone-2 work. You think the pace is faster at the same heart rate but you're not sure. You ask the agent to draw the curve — average pace, distance per session, weekly volume — and read what's actually happening underneath the feel.

What the agent needs to hear

Behind your sentence, the agent works backwards to three pieces: which activity, the window, and which metric. The activity can be a single discipline — running, cycling, rowing — or « all cardio » when you want the combined conditioning picture. The window defaults to the last three months. The metric defaults to pace plus distance per session, with weekly volume sitting underneath.

« How’s my running been » carries the activity and lets the defaults handle the rest. If you want the all-up view, name it: « all cardio, last 3 months ». Multi-activity views overlay the curves on the same card so you can read them against each other.

Activity, window, and metric choices

A single-activity view is cleaner for decisions — « is my running getting faster » reads better off a run-only curve than off a mixed one. « All cardio » is the right call when you want overall conditioning load rather than per-discipline progression.

The window defaults to three months. Override with named ranges (« since the cut started »), relative phrases (« last 8 weeks »), or explicit dates (« since March 1 »). Pick the window that matches the question you’re holding.

Metric options run wider than the default. Average pace per session is the gut-check on whether you’re getting faster. Distance per session shows whether your typical effort is growing. Weekly cardio volume — in minutes or kilometers — answers the base-building question. Average heart rate, if you log it, anchors the others. Ask for one or several; the curves render side by side.

What the curve actually tells you

The cardio curve reads three things. Endurance progression — pace climbing at the same heart rate, or heart rate dropping at the same pace — means the engine is improving. That’s the read most users actually want and rarely ask for directly. Volume trend — weekly minutes or kilometers going up — means base building is happening. Session shape drift is subtler: if the average distance per session shrinks but the session count holds, you’re doing more shorter sessions, which is a different stimulus than steady long ones.

Pair this read with the weight curve when relevant. Dropping weight while holding cardio pace usually means the strength side is intact. Dropping weight while pace also drops means the engine is going with the weight — a signal to eat more or back off the deficit.

When the agent gets it wrong

Three failure modes show up most. The activity got bucketed wrong — a hike landed in the run column, pulling the run-pace average down: « exclude the May 3 entry from the run trend, it was a hike ». Pace looks inconsistent because some sessions logged duration only and no distance: « show only sessions with distance ». And if the window crosses a deload week, the visual trend flattens for reasons that aren’t physiological: « show the curve excluding the deload week ».

What makes the cardio trend worth trusting

Three things decide whether this cardio curve tells you something useful: activity buckets are clean (runs in the run column, bike rides in the bike column — a mixed average is unreadable), the cadence matches the question (weekly is right for volume, per-session for pace, daily is too noisy for cardio), and the metric you read answers the question you’re asking (pace for fitness, volume for base, heart rate for effort). Cardio trends paired with strength and weight reads give you the full conditioning picture in one glance.