invoice-organizer
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (PDF and image content) to extract information. If these files contain hidden or malicious instructions, the agent could potentially be misled into performing unintended actions.
- Ingestion points: SKILL.md (Step 2: Extract Information from Each File) identifies file scanning and content extraction from PDFs and images as primary inputs.
- Boundary markers: Absent. The instructions do not specify using delimiters or 'ignore embedded instructions' warnings when reading document content.
- Capability inventory: find, mkdir, cp, mv (SKILL.md).
- Sanitization: Partial. While it suggests removing special characters from filenames, it does not validate the content extracted from documents before processing.
- [COMMAND_EXECUTION]: The skill utilizes shell commands to perform its core functions.
- Evidence: Use of find, mkdir, cp, and mv within SKILL.md to manage files and directories.
Audit Metadata