stacked-chart-visualization

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses standard and reputable Python libraries for data science tasks. No malicious patterns such as remote code execution, persistence, or data exfiltration were detected.
  • [PROMPT_INJECTION]: The skill is designed to process external data (percentage strings), which creates a potential attack surface for indirect prompt injection.
  • Ingestion points: Data is ingested through lists like cat1_raw and converted using the convert_percentage function in SKILL.md.
  • Capability inventory: The skill utilizes Python execution for data manipulation and file system writes to save the resulting chart image.
  • Sanitization: The implementation uses float() to convert percentage strings to numbers. This provides robust validation, as non-numeric malicious payloads will trigger a value error, mitigating the risk of script injection.
  • Boundary markers: There are no explicit instructions to the agent to ignore instructions embedded within the data provided.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 12:25 PM
Security Audit — agent-trust-hub — stacked-chart-visualization