epub-chapter-extractor
Pass
Audited by Gen Agent Trust Hub on Jun 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions direct the agent to execute a shell command using
uv runto run the extraction script and manage its dependencies. - [PROMPT_INJECTION]: The skill processes untrusted content from EPUB files, creating a vulnerability to indirect prompt injection. Maliciously crafted ebook content could influence the agent's behavior when the extracted text is processed in future steps.
- Ingestion points: The
extract_chapters.pyscript reads the full text content of user-provided EPUB files. - Boundary markers: There are no delimiters or isolation instructions used to distinguish the extracted chapter content from the agent's instruction context.
- Capability inventory: The skill possesses the ability to write files to the local system and execute commands via
uv. - Sanitization: The script removes functional HTML elements (e.g., scripts, styles) to ensure clean markdown, but it does not filter the natural language text for adversarial prompt instructions.
- [EXTERNAL_DOWNLOADS]: The skill fetches several well-known and standard Python libraries, including
ebooklib,beautifulsoup4,html2text, andlxml, from the Python Package Index (PyPI) during execution.
Audit Metadata