media-use

Managing media assets in a single command-line interface allows you to resolve, generate, and transform audio, image, and video files deterministically. It uses a global cache and provider catalog to ensure content-addressed reuse and metadata tracking across multiple project environments.

Automating background music resolution and voiceover generation involves using a unified interface to fetch matching tracks and generate audio via integrated local models. It keeps the project directory clean while registering all fetched assets in a central manifest.

Performing audio ducking and transcript-driven cutting is supported natively without manual context bloat. The skill executes these professional-grade media operations directly through its command-line interface, ensuring deterministic and repeatable results.

This media management skill works directly with local models and dependencies like ffmpeg, mflux, and parakeet-mlx. It integrates these external providers and local models to handle automated generation, catalog retrieval, and local transformation tasks.

The best way to apply cinematic color grading to video assets locally is through a unified command-line system that resolves color grades from a provider catalog. This ensures consistent content-addressed asset reuse and accurate metadata tracking.

Resolving diverse media assets across projects leads to fragmentation when using disconnected tools and manual workflows. This skill solves that by providing a single, deterministic interface for resolving, generating, and operating on audio, image, and video assets.