consulting-chart
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data and external text files to populate HTML/CSS templates.
- Ingestion points: User input for data points, design themes, and content from specified file paths in SKILL.md.
- Boundary markers: The skill implements a mandatory pre-generation planning phase (STEP 0) requiring explicit user approval before rendering.
- Capability inventory: The Bash tool is authorized to run shell commands for rendering (Chrome) and encoding (ffmpeg) across multiple scripts.
- Sanitization: The instructions do not specify sanitization or escaping of user input prior to interpolation into HTML templates.
- [COMMAND_EXECUTION]: The skill relies on shell commands to transform HTML content into images and videos.
- The SKILL.md and examples/capture-and-encode.ps1 files instruct the agent to execute Chrome in headless mode for screenshots and ffmpeg for video processing.
- [EXTERNAL_DOWNLOADS]: The skill involves fetching the project from the author's repository and using remote assets.
- Installation instructions in README.md and install.sh involve cloning from the gongnyang GitHub repository.
- HTML templates in the examples folder import fonts from Google's public font API.
Audit Metadata