pdf-extractor
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill extracts text, tables, and metadata from PDF files. If a user processes an untrusted PDF containing hidden instructions (e.g., in white text, metadata fields, or small fonts), these instructions will be converted to plain text and fed back into the agent's context. This could influence or override the agent's subsequent behavior. The risk is classified as LOW as it depends on the content of the processed files rather than malicious code within the skill itself.
Audit Metadata