scroll-stop-builder
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
ffprobeandffmpegcommands to analyze video metadata and extract image frames for high-performance canvas rendering. - [COMMAND_EXECUTION]: It uses
python3 -m http.serverto launch a local development server, enabling the user to preview the generated site in a browser. - [EXTERNAL_DOWNLOADS]: The generated website template includes the Three.js library from Cloudflare's public CDN (
cdnjs.cloudflare.com) to support optional 3D particle effects. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it is designed to ingest data from user-provided URLs via the
WebFetchtool (documented in the Step 0 Interview inSKILL.md). 1. Ingestion point: Web content retrieved during the branding interview. 2. Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard instructions within the fetched data. 3. Capability inventory: The agent can execute shell commands (ffmpeg) and perform file system operations. 4. Sanitization: No sanitization or filtering of the external content is specified.
Audit Metadata