WorkCapacity Blog
Ideas for athletes, coaches, and systems that optimize real output.
Read articles on measurement, power, adaptation, and the logic behind AI-native coaching. This is where WorkCapacity explains the category, sharpens the thesis, and publishes ideas worth sharing.
Power Oracle now computes fitness across broad time and modal domains
New `fitness_score`, `modality_curves`, and `coverage_matrix` fields on the athlete curve endpoint. The geometric mean of per-modality work-capacity AUCs, sampled via cumulative chains.
Recent Posts
Clear writing on the signals that actually matter
Power Oracle now fully supports L402 payments
Clients can now pay for Power Oracle compute with L402 in production, with the same public pricing contract as x402 and a cleaner path for Lightning-native integrations.
Why CrossFit workouts are hard to model
Mixed workout formats break naive tracking. If the system flattens structure too early, athletes lose explanation and agents lose a reliable object to optimize.
What AI agents need from fitness metrics
If the user is an optimization system instead of a person, proxy metrics stop being good enough. The measurement object has to be computable.