data-import-parsers

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • NO_CODE (SAFE): The skill is entirely instructional markdown (SKILL.md) and contains no executable scripts, configuration files, or command-line instructions.
  • General Security (SAFE): The instructions promote secure engineering practices, such as explicit type validation, memory-safe streaming, and idempotent database operations, which are beneficial for system stability and security.
  • Indirect Prompt Injection (SAFE): While the skill discusses processing external data (CSV/JSON), it provides a framework for engineering the parser logic rather than defining prompt templates that would be vulnerable to indirect injection. The emphasis on strict type coercion and validation acts as a defensive measure.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 11:21 AM