story-copyright-detector

Franchise IP risk in story seeds is detected by analyzing text for known franchises or franchise-like artifacts. The detector outputs a JSON object with ip_level, franchise, model_preference, and reason to drive compliant model routing decisions.

IP-safe seed classification categorizes story inputs to determine rendering routes before image generation. You need it at canvas-enter time to route seeds with potential IP constraints to compliant pipelines and avoid copyright violations.

Route image generation by reading the detector's JSON output containing ip_level and model_preference. Use these fields to direct seeds with high IP risk to compliant models while sending generic archetypes to safe default pipelines.

Yes, you can use it for automated content moderation by providing a seed input and reading the exact JSON routing decision. It outputs exactly one object containing ip_level, franchise, model_preference, and reason for downstream policy routing.

Automated franchise detection limitations include relying on identifying known franchises or franchise-like tokens in the seed text. It may classify generic archetypes as safe defaults, but complex or obscure IP references might not trigger the correct model_preference routing.

Yes, the story copyright detector works with existing image pipelines without requiring external dependencies. It functions independently to classify seeds and outputs a compact JSON payload for seamless integration into your downstream routing logic.