baoyu-url-to-markdown

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill fetches and converts external web content, which can contain instructions intended to manipulate the AI agent.\n
  • Ingestion points: scripts/main.ts fetches arbitrary web content from user-provided URLs.\n
  • Boundary markers: The skill uses YAML frontmatter to separate metadata, but the converted markdown body lacks explicit delimiters or warnings to ignore embedded instructions.\n
  • Capability inventory: The agent context using this skill has capabilities for file system writing (writeFile, mkdir) and local script execution.\n
  • Sanitization: scripts/html-to-markdown.ts removes executable HTML tags (script, iframe, etc.), but does not filter or sanitize text-based natural language instructions.\n- [Dynamic Execution] (LOW): The skill uses evaluateScript to execute a static JavaScript template for page cleaning. While this is dynamic execution, the script is hardcoded and does not incorporate untrusted input into the code logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:28 PM