ie-engine

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill functions as a logical controller for internal note-processing sub-skills and does not perform unauthorized actions or external communications.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes historical user notes, creating a theoretical surface for indirect prompt injection, although this is inherent to its primary purpose of analyzing past data.
  • Ingestion points: Historical note data processed via the memory layer in SKILL.md and references/pipeline.md.
  • Boundary markers: No explicit content delimiters are defined in the orchestration logic.
  • Capability inventory: Orchestrates reasoning and text generation through internal sub-skills.
  • Sanitization: No content validation or escaping is described in the pipeline definitions.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 04:00 AM