sheltered-instruction-lesson-modifier

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it ingests untrusted data that is then interpolated into the model's instructions.
  • Ingestion points: The fields lesson_plan, student_profiles, eal_students, and resources_available in SKILL.md take external input.
  • Boundary markers: Absent. The prompt uses simple labels (e.g., **Lesson plan:** {{lesson_plan}}) without formal delimiters or explicit instructions to the agent to disregard commands within the input.
  • Capability inventory: The skill is restricted to generating text based on the output_schema. There are no subprocess calls, file system operations, or network requests defined within the skill.
  • Sanitization: Absent. There is no evidence of input validation or escaping before the data is processed by the model.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 10:53 PM