nanobanan-ppt
Warn
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill's installation process involves cloning a repository from a community source (
https://github.com/op7418/NanoBanana-PPT-Skills.git).- [COMMAND_EXECUTION]: The installation steps require executing shell commands to clone the repository, create a Python virtual environment, and install dependencies via pip.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input documents to generate slide content and structures, creating a vulnerability surface for indirect instructions to influence agent behavior. - Ingestion points: Processes external documents provided by the user for content and tone analysis.
- Boundary markers: No specific delimiters or safety instructions are mentioned for separating document content from slide generation logic.
- Capability inventory: Generates images via Gemini API, generates videos via Kling AI API, and writes output to slide images and an HTML player.
- Sanitization: No sanitization or validation of the input document content is described.
Audit Metadata