production-validator

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill is designed to ingest and process XML data from an external source ('screenwriter' skill), creating a surface for indirect prompt injection.
  • Ingestion points: XML input received from the screenwriter skill as defined in SKILL.md.
  • Boundary markers: The skill relies on XML structure but lacks explicit instructions to ignore potentially malicious prompts embedded within the scene action or metadata tags.
  • Capability inventory: The skill generates output prompts that are passed to downstream skills ('imagine', 'arch-v'), meaning a malicious injection in the input could influence the behavior of subsequent agents.
  • Sanitization: There is no evidence of sanitization or filtering logic for the input XML before it is used to construct prompts.
  • [No Code] (SAFE): No executable scripts (Python, JavaScript, Shell) are included in the skill. All logic is described through natural language instructions and reference documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:49 PM