Build a training dashboard from connected run history

Why use this

1. A runner may have many recorded sessions but struggle to see which periods include consistent activity and which contain missing or duplicated uploads. A dashboard can look persuasive even when different services describe the same run differently.

2. This job could organize an approved history into a transparent training view, with data gaps and definitions visible. It can support personal reflection while keeping physiological conclusions and training prescriptions outside what the records alone establish.

Why an agent / simpler alternative

Useful judgment

Reconcile session records, distinguish missing data from inactivity and explain comparison limits.

Simpler alternative

Use the running service's built-in charts when they already answer the question.

Unsuitable when

The dashboard would diagnose injury, prescribe training or require access beyond the runner's consent.

Proposed inputs

1. Name

   Run export

   Owner

   Runner

   Minimum scope

   Chosen period of activity dates, durations and distances

   Quality check

   Check units, duplicates and timezone

2. Name

   Dashboard brief

   Owner

   Runner

   Minimum scope

   Questions and preferred aggregation periods

   Quality check

   Separate descriptive trends from health interpretations

Access and setup — proposed or unknown

Proposed minimum

1. Start from a supplied export with location traces removed; no new account or fallback connection.

Compatibility unknowns

1. Current service access, export coverage and runtime compatibility are unknown; preserve the referenced access failure.

External actions requiring approval

1. Connecting another service, creating accounts, sharing health-related records or publishing the dashboard requires explicit approval.

Expected deliverable and review checks

Proposed artifact

Draft activity dashboard with metric definitions, data-quality flags and source-period coverage.

Critical checks

1. Reconcile totals to the export; distinguish zero activity from absent data; keep distance units consistent.

Missing output evidence

This overlay adds no new output inspection, execution or validation; consult the referenced canonical source outcome limits.

Failure signals and proposed responses

1. Signal

   One run appears in two imports

   Why it matters

   Totals exaggerate training volume

   Proposed response

   Mark probable duplicates and request confirmation before exclusion.

2. Signal

   A service outage looks like a rest week

   Why it matters

   The chart encourages an unsupported interpretation

   Proposed response

   Show a coverage gap rather than a zero.

Effort and cost — unknown

Setup work

1. Agree export scope and metric definitions.

Ongoing review

1. Review duplicates, gaps and interpretation boundaries.

Unknowns

1. Export completeness, historical service behavior, dashboard upkeep and costs are unknown.

First check — proposed, unexecuted

Status

proposed_unexecuted

Synthetic fixture

Invent six runs with mixed distance units, one duplicate and a missing-data interval; sketch a weekly summary.

Acceptance checks

1. Normalize units; show the duplicate separately; label the gap; make no health recommendation.

Stop conditions

1. Stop if work needs an unapproved service connection or sensitive location disclosure.

Limits

1. Descriptive running history does not establish fitness, recovery needs or a safe training load.