excel-variance-analyzer

Pass

Audited by Gen Agent Trust Hub on Mar 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted financial data from external sources (Excel, CSV) that could contain malicious instructions.
  • Ingestion points: Budget and actual data files (Excel, CSV, or pasted tables) as specified in Step 1 of SKILL.md.
  • Boundary markers: Absent; the instructions do not require the use of delimiters or specific warnings to ignore embedded commands when processing external content.
  • Capability inventory: Significant tools are available, including file system access (Read, Write, Edit, Glob, Grep) and shell command execution via Bash (npx).
  • Sanitization: Absent; while the skill checks for numeric values, it lacks validation to filter out or escape instructional text that might be hidden in data headers or non-numeric fields.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 28, 2026, 01:42 AM