variation-theory-task-designer

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill incorporates untrusted user data into its internal prompt sequence, creating a potential for indirect prompt injection.
  • Ingestion points: Untrusted data enters the agent context through the object_of_learning, common_confusion, student_level, subject_area, current_task, and lesson_context fields defined in SKILL.md.
  • Boundary markers: Absent. The skill does not use delimiters to isolate user inputs or instructions for the agent to ignore embedded directives within that data.
  • Capability inventory: The skill is limited to generating pedagogical text and does not utilize any capabilities for file system modification, network access, or shell command execution.
  • Sanitization: No input validation or escaping is applied to the user-provided strings before they are inserted into the prompt template.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 10:54 PM