The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for executing various CLI commands via npx hyperframes, including project initialization, linting, previewing, and rendering. These are standard development workflows for the framework.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx, which downloads and executes the latest version of the hyperframes package from the npm registry. It also references the official HyperFrames GitHub repository (heygen-com/hyperframes) for source and documentation.
[INDIRECT_PROMPT_INJECTION]: The skill processes external project files (HTML, JavaScript, and media) during linting and rendering. There is a theoretical risk that malicious content within these files could attempt to influence the agent's behavior if the tool's output is consumed by the agent without sanitization.
Ingestion points: Files in ./, compositions/, and media files like clip.mp4 or track.mp3.
Boundary markers: Not explicitly defined in the CLI instructions.
Capability inventory: File system access, network access (via npm/transcription), and subprocess execution (via FFmpeg/Chrome).
Sanitization: Not documented; relies on the internal security of the HyperFrames CLI tool.

hyperframes-cli