semantius-modeler

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). In semantius-modeler Stage 1/Step 1 runtime, the agent reads the user-provided *-semantic-spec.md file produced by semantius-analyst (“Locate the *-semantic-spec.md file”), then parses multiple free-text fields (descriptions, JsonLogic message/description, enum labels, and Select rule/Validation rules objects) from that file into the model and uses them to execute Semantius writes via Stage 4.