# V2 Nous Anti-Cheat Classifiers

`@v2/nous-anti-cheat-classifiers` binds V2 fair-play telemetry to the Nous
anti-cheat classifier suite.

- V2 publishes labeled training examples on `@oshun/event-bus` topics.
- `@nous/training` owns the training loop for smurf, win-trading,
  coordinated-throw, and geographic-anomaly classifiers.
- `@nous/safety` hosts the Model Card metadata at
  `V2/docs/ai/model-cards/anti-cheat-classifier-suite.md`.
- Classifier output is off-rollback, review-only, and cannot issue automatic
  account discipline.

The bridge returns the training plan and Model Card together so release gates
can assert classifier coverage, event-bus topic coverage, human review, appeal
coverage, and rollback isolation in one place.