instagram-tweet-carousel
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the 'playwright' package and its associated Chromium browser via 'pip3' and 'playwright install'. These are well-known tools used for the legitimate purpose of exporting HTML slides to PNG images.
- [COMMAND_EXECUTION]: Utilizes system commands like 'file' to determine image MIME types and 'pip3' for dependency management, which are routine operations for processing image assets and setting up the environment.
- [DYNAMIC_EXECUTION]: Dynamically generates and executes a Python script that leverages Playwright to automate browser rendering and slide screenshotting. It also performs JavaScript execution within the browser context via Playwright's 'evaluate' function to modify the layout for export.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided text for slide content and branding details and embeds them into an HTML template for rendering, creating a potential injection surface.
- Ingestion points: User-provided topic, display name, handle, and content processed in SKILL.md.
- Boundary markers: None; the skill does not use delimiters or specific instructions to ignore embedded commands in user-supplied strings.
- Capability inventory: Subprocess execution (shell, Python, Playwright), file-system writes, and network operations (Google Fonts) as described in the SKILL.md instructions.
- Sanitization: None; there are no instructions to sanitize, escape, or validate user-supplied text before it is interpolated into the HTML structure.
Audit Metadata