wan-pptx-generator

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a Python script (scripts/pptx-generator.py) to manage the end-to-end workflow of parsing inputs, interacting with the DashScope API, and downloading output assets.
  • [EXTERNAL_DOWNLOADS]: The script automatically downloads slide images and PowerPoint files using urllib.request.urlretrieve. While intended for Aliyun-hosted assets, the regular expression used to identify the PPTX download link in scripts/pptx-generator.py does not validate the domain, potentially allowing the download of files from arbitrary external sources if the AI model output is manipulated.
  • [DATA_EXFILTRATION]: To enable document-to-PPT conversion, the skill uploads user-provided files (PDF, DOCX, TXT, etc.) to Aliyun's DashScope infrastructure. This is a primary function but involves transmitting local document data to an external service.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from user-provided documents, which could contain malicious instructions designed to influence the AI's behavior or output.
  • Ingestion points: Files are ingested via the -f or --file command-line arguments and read or uploaded in scripts/pptx-generator.py.
  • Boundary markers: The skill does not implement delimiters or specific instructions to isolate the ingested document content from the model's system instructions.
  • Capability inventory: The skill has capabilities for network uploads/downloads and local file system writes within the designated output directory.
  • Sanitization: No content filtering or sanitization is performed on the document data before it is processed by the AI model.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 07:37 AM