skills/liustack/pagepress/markpress/Gen Agent Trust Hub

markpress

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @liustack/markpress package from the NPM registry and uses npx playwright install chromium to download the Chromium browser binary. These resources originate from the skill's author and well-known service providers.
  • [COMMAND_EXECUTION]: Executes the markpress CLI tool for document conversion and the npx playwright command for environment initialization.
  • [PROMPT_INJECTION]: The skill processes external Markdown data, which constitutes an indirect prompt injection surface.
  • Ingestion points: Reads user-provided Markdown files via the -i parameter as described in SKILL.md.
  • Boundary markers: The conversion pipeline includes rehypeSanitizeTags, which serves to strip or modify incoming content to meet safety constraints.
  • Capability inventory: File system write operations (-o), system clipboard access (--copy), and dynamic browser-based rendering of diagrams via Playwright.
  • Sanitization: Explicitly removes dangerous tags such as <script>, <style>, and <iframe>, and downgrades external links to text-based footnotes to prevent execution of malicious code in the rendering environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 11:51 AM
Security Audit — agent-trust-hub — markpress