doc-to-vector-dataset-generator
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external, untrusted document content (PDF, Markdown), creating a potential attack surface for indirect prompt injection.
- Ingestion points: Text is extracted from local files using the
extract_pdfandextract_markdownfunctions withinSKILL.md. - Boundary markers: The skill does not implement boundary markers or instructions to downstream systems to ignore potential commands embedded in the processed text chunks.
- Capability inventory: The skill possesses file system read capabilities via
openandglob, and write capabilities via theexport_jsonlfunction. - Sanitization: While the
clean_textfunction performs noise reduction (removing whitespace and page numbers), it lacks logic to identify or sanitize malicious instructions embedded in the document text.
Audit Metadata