The Agent Skills Directory

[SAFE]: The skill incorporates rigorous safety guidelines and boundaries. It specifically instructs the agent to use fictional demo data and deterministic environments, explicitly forbidding the use of production credentials, real user data, or permanently mutating flows during recording sessions.
[COMMAND_EXECUTION]: The skill provides comprehensive documentation and templates for utilizing standard tools such as Playwright and ffmpeg. These command examples are used for legitimate media production tasks (recording, transcoding, and quality analysis) and are intended for use within controlled development or CI environments.
[EXTERNAL_DOWNLOADS]: The documentation references several well-known and trusted service providers (e.g., OpenAI, Microsoft, ElevenLabs) for specialized capabilities like text-to-speech and automated transcription. These references are standard for the skill's stated purpose and do not involve untrusted third-party code execution.
[DYNAMIC_EXECUTION]: The skill uses Playwright's ability to evaluate JavaScript within a browser context to generate visual effects like click ripples and overlays. These implementation patterns are standard for high-quality demo recording and are used here to enhance the narrative quality of the output in a safe, sandboxed browser environment.

cue