The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/project_setup.py utility uses subprocess.run to execute npm install within the generated project directory. This is an intended feature to facilitate project initialization by installing the required Barba.js, GSAP, and Vite dependencies.
[INDIRECT_PROMPT_INJECTION]: The skill documentation describes methods for ingesting untrusted data and updating the application state. 1. Ingestion points: Data enters the agent's context through AJAX-fetched HTML containers and page metadata as shown in references/hooks_guide.md and references/api_reference.md. 2. Boundary markers: The provided examples do not include explicit boundary markers or instructions to ignore embedded content in the fetched HTML. 3. Capability inventory: The skill includes file system writes and command execution capabilities via scripts/project_setup.py and scripts/transition_generator.py. 4. Sanitization: The examples demonstrate direct interpolation of page content (e.g., document.title = next.html.querySelector('title').textContent) without sanitization or validation.

barba-js