epub-to-markdown

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill performs legitimate file format conversion.\n- [COMMAND_EXECUTION]: The skill consists of a standalone Python script intended for local file processing. It does not invoke shell commands, subprocesses, or dynamic code execution (eval/exec).\n- [DATA_EXFILTRATION]: The skill does not contain any code for network operations (e.g., requests, urllib.request, curl). It reads local EPUB files and writes Markdown output locally.\n- [PROMPT_INJECTION]: The skill processes untrusted data from EPUB files, which is a surface for indirect prompt injection. However, it does not interpret the content as instructions for the agent.\n
  • Ingestion points: Reads content and metadata from the input EPUB file via epub_to_markdown.py.\n
  • Boundary markers: The output Markdown uses structured headings and a YAML-like metadata block, providing basic structural separation.\n
  • Capability inventory: Limited to local file read and write operations; no network or administrative capabilities are present.\n
  • Sanitization: Uses BeautifulSoup to strip scripts and styles from the source HTML and html2text for clean markdown conversion.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 02:49 PM