The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocesses to manage and generate media files.\n
The script download_videos_from_manifest_with_ytdlp.mjs invokes python3 -m yt_dlp to download video content from provided manifests.\n
The workspace runtime in postplus_workspace_runtime.mjs utilizes ffmpeg for generating test video fixtures and uses the macOS open utility to allow users to view project files.\n- [EXTERNAL_DOWNLOADS]: The skill contains functionality to download external media and interact with cloud APIs.\n
Video assets are downloaded from remote sources via yt-dlp based on user-supplied manifests.\n
The network_runtime.mjs and postplus_cloud_client.mjs libraries facilitate communication with PostPlus Cloud APIs for authentication and media processing.\n- [PROMPT_INJECTION]: The skill processes structured data that could serve as an indirect prompt injection vector.\n
Ingestion points: The skill reads user-provided JSON files such as brief.json and manifest.json in build_video_request_architecture.mjs and download_videos_from_manifest_with_ytdlp.mjs.\n
Boundary markers: Instructions in SKILL.md advise making requests self-contained and avoiding references to prior segments.\n
Capability inventory: The skill has access to the local file system, network requests, and subprocess execution (ffmpeg, python3, open).\n
Sanitization: File operations are restricted to the project workspace through path resolution logic in postplus_workspace_runtime.mjs.

video-request-architect