xhs-note-creator

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires downloading Chromium browser binaries through the Playwright library to perform image rendering. These downloads target official project repositories and are standard for the tool's purpose.
  • [COMMAND_EXECUTION]: The rendering scripts utilize Playwright to launch a headless browser and execute internal JavaScript routines to calculate content height and apply visual styles during image generation.
  • [DATA_EXFILTRATION]: The skill manages Xiaohongshu session cookies provided by the user in a local environment file (.env) for note publication. It includes an optional mode that can send these credentials to a user-defined endpoint, which defaults to localhost.
  • [SAFE]: The skill possesses an indirect prompt injection surface (Category 8) as it processes user-supplied data for rendering. Ingestion points: Markdown files provided to scripts. Boundary markers: None present. Capability inventory: Playwright browser execution for image rendering. Sanitization: The skill uses the standard 'markdown' (Python) and 'marked' (Node.js) libraries for conversion. The workflow is typical for local developer tools and no malicious intent was found.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 03:37 AM
Security Audit — agent-trust-hub — xhs-note-creator