document-converter-suite
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external files such as PDFs, Word documents, and HTML pages. Maliciously crafted documents could contain hidden instructions intended to override the agent's behavior when the extracted text is subsequently processed in the agent's context.
- Ingestion points: Files are read and processed in
scripts/convert.py,scripts/batch_convert.py,scripts/table_extractor.py, andscripts/form_filler.py. - Boundary markers: Absent. The conversion process extracts text and structural elements without wrapping them in security delimiters or providing explicit instructions to the agent to ignore any embedded commands within the content.
- Capability inventory: The skill possesses capabilities for reading and writing to the local file system across several scripts and libraries. It does not perform network operations.
- Sanitization: While the skill includes basic escaping for HTML (
scripts/lib/html_writer.py) and PDF (scripts/lib/pdf_writer.py) output, it does not perform semantic sanitization of the extracted text content to mitigate the risk of prompt injection attacks.
Audit Metadata