data-analysis

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses standard office automation APIs to create reports and perform calculations. All code snippets are transparent and restricted to the active workbook.
  • [PROMPT_INJECTION]: The skill processes spreadsheet data to generate natural language conclusions, which creates a surface for indirect prompt injection. This is evaluated as safe as it is fundamental to the skill's primary purpose and does not involve high-privilege capabilities.
  • Ingestion points: Untrusted cell data is ingested via srcWs.Range(...).Value2 in SKILL.md during pivot calculations and conclusion generation.
  • Boundary markers: Absent; the AI is instructed to generate conclusions directly from data values without specific output delimiters.
  • Capability inventory: Limited to spreadsheet manipulation (reading/writing cells and formulas); no network access or system command execution capabilities are requested or used.
  • Sanitization: Not present; the skill relies on the spreadsheet engine to interpret values and the AI to interpret data for conclusions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 07:32 AM
Security Audit — agent-trust-hub — data-analysis