The Agent Skills Directory

[SAFE]: The skill's operations are consistent with its stated purpose as a media production and image generation tool for video professionals.
[COMMAND_EXECUTION]: Executes system commands for ffmpeg and python3 -m yt_dlp to perform media processing and video downloads within the project workspace. These calls are used to generate test patterns and download reference media.
[DATA_EXPOSURE]: Reads local configuration files from standard OS locations (e.g., ~/Library/Application Support/postplus) to manage session tokens for PostPlus Cloud API authentication. This is standard behavior for vendor-provided CLI tools.
[INDIRECT_PROMPT_INJECTION]: Manages an attack surface for indirect prompt injection by processing external prompt data through its workspace dashboard and JSON request files.
Ingestion points: postplus_workspace.mjs (web-based asset management) and scripts/generate_image.mjs (CLI-driven request processing).
Boundary markers: No explicit markers are used to isolate prompt content in API requests.
Capability inventory: Network access for API calls, file system writes for asset storage, and subprocess execution for media tools.
Sanitization: Validates asset identifiers and employs strict path resolution logic to ensure file operations remain within the designated project directory.

image-batch-runner