Skip to main content
Version: Latest (unreleased)

Static

Overview​

static provides deterministic model choice without metrics or learned state. It selects the first entry in a decision's modelRefs by default. A matched domain can supply fixed model_scores; the selector uses the highest score when at least one candidate has a configured score. Explicit scores of 1.0 participate in ranking like any other configured score.

Key Advantages​

  • Deterministic and easy to audit.
  • Has no selector model, metrics, or storage dependency.
  • Provides a stable baseline for other selection policies.

What Problem Does It Solve?​

Some routes already have an intentional model order or fixed per-domain scores and do not need an online ranking policy. Static makes that choice explicit.

When to Use​

Use Static for deterministic routing, as a baseline when comparing selectors, or when an external process owns candidate ordering. If a decision has only one candidate, you can usually omit the algorithm entirely.

Configuration​

algorithm:
  type: static

Place the intended fallback winner first in modelRefs. To rank with domain model_scores, score every candidate. Unscored candidates retain the default score of 1.0; if no candidate has a configured score, the selector falls back to the first candidate. Equal highest scores also keep the first candidate in the tied set. See a complete example: config/fragments/algorithm/selection/static.yaml.

Dependencies and Limitations​

  • No external dependencies and no request-content processing beyond ordinary decision matching.
  • It does not fail over to later candidates, react to load, or learn from outcomes. Backend availability remains the responsibility of the normal provider path.