The Agent Skills Directory

[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by processing content from arbitrary user-provided URLs.
Ingestion points: Data enters the agent's context through capture/extracted/visible-text.txt and capture/extracted/asset-descriptions.md after the hyperframes capture command is executed.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands within the ingested text are specified in the provided reference files.
Capability inventory: The skill has the capability to execute shell commands (npx hyperframes), write files to the local system (compositions/*.html), and perform network operations via external TTS and Vision APIs.
Sanitization: The skill lacks explicit sanitization or filtering of the extracted website text before it is used to generate the storyboard and narration script.
[COMMAND_EXECUTION]: The skill relies on executing various shell commands using npx hyperframes for project initialization, site capture, linting, validation, and rendering. These commands run with the permissions of the user environment.
[EXTERNAL_DOWNLOADS]: The skill facilitates data transmission to and from external services including ElevenLabs, HeyGen, and Google Gemini for text-to-speech, transcription, and image analysis. These are well-known services used for the skill's primary functionality.

website-to-video