creative-tools
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: The skill processes user-provided files (PDFs and text) for analysis and transformation. This is the primary function of the skill and involves standard file reading operations without unauthorized access to sensitive system directories.
- [EXTERNAL_DOWNLOADS]: The documentation references various external AI services (Google Gemini, ElevenLabs, OpenAI, Flux, etc.) for image and voice generation. These are presented as optional integrations for the user and do not involve automated execution of untrusted remote code.
- [COMMAND_EXECUTION]: The PDF processing module utilizes standard command-line tools like qpdf and pdftk. The instructions guide the user/agent on how to use these tools for legitimate document manipulation tasks.
- [INDIRECT_PROMPT_INJECTION]: As the skill ingests external data (PDFs and blog content), it has a surface for indirect prompt injection. However, the skill includes rigorous multi-step validation protocols (e.g., automated bounding box checks and manual visual inspection in pdf-processing__references/forms.md) to mitigate risks associated with processing untrusted content.
- [DYNAMIC_EXECUTION]: The script pdf-processing__scripts/fill_fillable_fields.py contains a runtime monkeypatch for the pypdf library. This is a targeted fix for a known issue in the library's form-handling logic and does not represent a malicious dynamic execution vector.
Audit Metadata