The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs several shell scripts (setup.sh, gen-audio.sh, install.sh) to handle the installation of the skill into various agent environments and to automate the video project setup. These actions are consistent with the skill's primary purpose of creating a local development environment for video rendering.
[EXTERNAL_DOWNLOADS]: During setup, the skill performs an npm install to fetch standard React and Remotion libraries. It also optionally downloads audio assets from well-known AI services (ElevenLabs and OpenAI) to provide voiceovers for the generated videos.
[CREDENTIALS_UNSAFE]: Authentication for third-party services is managed through standard environment variables (ELEVENLABS_API_KEY, OPENAI_API_KEY). The skill correctly instructs users to provide these keys in their environment rather than hardcoding them in the project files.
[DATA_EXFILTRATION]: Narration text derived from user transcripts is sent to OpenAI or ElevenLabs for text-to-speech processing. This data transmission is documented as a core feature and targets official, well-known service endpoints.

meeting-to-video