Verifies the article's factual claims against the live web before scoring. Anthropic uses the model's built-in web search (Claude only). Tavily fetches web snippets, then the chosen model synthesizes the verdicts.

Scores the article across the eight veracity categories and returns the FVI score. The provider determines the API endpoint, so the model family must match it.

Each member scores the article independently and the per-category median becomes the result — and the training label for PRUUF's own judge. Three members is the minimum that does real work: a median of two is just their midpoint, so a dissenting member is averaged in at full weight instead of being outvoted. Every member added multiplies analysis cost.

The overnight news agent uses the settings above by default. Enable this to run the nightly batch on different (e.g. cheaper) models without changing interactive analysis.