pdf-toolkit
Pass
Audited by Gen Agent Trust Hub on May 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on the execution of a local Python script,
scripts/pdf_auditor.py, to perform its core function of auditing PDF documents. This script is called via shell commands in the described workflows. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted data from PDF files.
- Ingestion points: The script
scripts/pdf_auditor.pyreads the binary content of user-provided PDF files and extracts metadata strings. - Boundary markers: Absent. The script outputs metadata fields directly without using delimiters or protective instructions for the agent consuming the output.
- Capability inventory: The agent executes
scripts/pdf_auditor.pyto obtain document statistics and metadata. - Sanitization: Absent. While the script decodes various PDF string formats, it does not sanitize the resulting text for malicious instructions or prompt injection patterns.
Audit Metadata