The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs several Node.js packages from the @helios-project namespace (@helios-project/core, @helios-project/renderer, @helios-project/cli) to facilitate video rendering and composition.
[COMMAND_EXECUTION]: The skill utilizes the npx helios render command to process the generated HTML composition into a final MP4 video file.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data by researching brand identity from either a user-provided URL or local codebase files such as documentation and CSS stylesheets.
Ingestion points: External marketing sites/URLs and local project assets (Tailwind configs, stylesheets, microcopy).
Boundary markers: There are no explicit instructions or delimiters used to isolate researched content from the agent's core instructions.
Capability inventory: The skill has the capability to install packages, write local HTML/JS files, and execute CLI tools.
Sanitization: The skill does not define specific sanitization routines for the data extracted during the brand research phase.
[CREDENTIALS_UNSAFE]: While the skill mentions the requirement of an ELEVEN_LABS_API_KEY, it follows security best practices by explicitly instructing the agent NOT to read .env files or verify keys itself, instead directing the user to manage the key in their environment.

guided-explainer-video