frontend-harness-slides
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as its primary function involves ingesting and transforming untrusted user-supplied content (e.g., transcripts, research memos, and document outlines) into React/Vite code and slide copy.
- Ingestion points: User-provided source material is processed during the 'Content Source' and 'Registry Draft' phases described in
references/01-plan.md. - Boundary markers: The instructions do not explicitly mandate the use of delimiters or 'ignore embedded instructions' warnings when the agent interpolates this untrusted content into its generation workflow.
- Capability inventory: The skill guides the agent to create filesystem artifacts and execute shell commands for building and testing the slide project.
- Sanitization: There are no specific instructions for sanitizing or escaping the content before transformation into code.
- [EXTERNAL_DOWNLOADS]: The skill fetches configuration and showcase images from the author's public GitHub repository (
patrick-fu/frontend-harness-slides-workbench) and references a live demonstration workbench hosted on Vercel. These resources are associated with the vendor and established cloud service providers.
Audit Metadata