code-data-analysis-scaffolds

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): Vulnerable to Indirect Prompt Injection (Category 8).
  • Ingestion points: User-provided datasets or codebases as described in SKILL.md (Step 1) and triggered by phrases like 'Explore this dataset'.
  • Boundary markers: Absent; the instructions do not include specific delimiters or warnings to ignore instructions embedded in the processed data.
  • Capability inventory: The skill involves generating and executing code using libraries such as pandas and pytest, as seen in SKILL.md (Step 5) and the examples.
  • Sanitization: Absent; no escaping or validation of external content is specified before interpolation into agent actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:09 PM