thenvsnow
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill includes instructions to run a local Python script (
scripts/validate.py) used for validating the generated HTML. This script uses the Playwright library to automate browser-based layout and functionality tests. - [EXTERNAL_DOWNLOADS]: The skill references Google Fonts (
fonts.googleapis.com), a well-known and trusted service, to provide typography for the optional handwritten page style. - [INDIRECT_PROMPT_INJECTION]: The skill accepts user-supplied topics and data to generate comparison pages, creating a potential surface for indirect injection.
- Ingestion points: User-provided topics and datasets processed through the
/thenvsnowcommand (SKILL.md). - Boundary markers: The instructions do not mandate the use of explicit delimiters or instructions to ignore embedded text in user-supplied data.
- Capability inventory: The agent utilizes web search and file-writing capabilities to produce and save the comparison pages (SKILL.md).
- Sanitization: The provided HTML template implements basic character escaping (
escfunction) for data-driven attributes to prevent injection in the generated output.
Audit Metadata