teaser-screenshots

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses several shell commands to manage the environment and run captures, including docker compose to list WordPress plugins and npm for building and running tests. These are standard operations for a development environment.
  • [EXTERNAL_DOWNLOADS]: The instructions suggest installing pngquant via Homebrew (brew install pngquant) to optimize images. This is a well-known service and a standard utility for this purpose.
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface: The skill automates UI interactions through Playwright, which involves ingesting data from a local WordPress installation.
  • Ingestion points: UI elements like user card popovers and meta lines are read by Playwright scripts.
  • Boundary markers: None identified for UI data.
  • Capability inventory: The skill can execute bash commands and write files to the assets/images/ directory.
  • Sanitization: No specific sanitization of UI content is mentioned before it is processed by the agent.
  • [SAFE]: The skill contains a hardcoded local path (/Users/bonny/Projects/...) belonging to the author. While this reveals the directory structure of the developer's machine, it does not expose sensitive credentials or pose a security risk in the context of a local development tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 01:21 PM
Security Audit — agent-trust-hub — teaser-screenshots