eval-config

Content evaluation thresholds are configured by setting minimum acceptable scores (0–100) per quality dimension, such as claim verification and hallucination risk, using the active brand profile to enforce brand governance rules.

Yes, you can adjust scoring weights to control how quality dimensions contribute to the composite score, and generate industry recommendations to align configuration settings with specific industry and brand priorities.

Auto-reject rules are established by defining a composite score floor that automatically fails substandard content, protecting quality and compliance by preventing low-scoring deliverables from passing quality gates.

Content-type overrides allow you to apply different quality bars for different formats, ensuring compliance-critical deliverables are evaluated with stricter scoring thresholds than awareness content.

Yes, the configuration process validates weight sums and threshold ranges to ensure mathematical accuracy, presenting effective scoring behavior with before and after impact analysis for your brand profile.

Run the recommend action to generate suggested configurations using brand context, industry priorities, and guideline requirements, providing industry-appropriate threshold and weight settings for your active brand profile.