This is a comprehensive forensic audio toolkit with 45 rules across signal preservation, noise profiling, spectral processing, and voice isolation. You'd reach for it when working with degraded recordings that need enhancement for transcription or legal evidence, or when authenticating audio for tampering. The standout piece is the workflow automation: it includes Python scripts for preflight analysis, workflow planning from metrics, and before-after comparison so you're not guessing whether your noise reduction actually helped. Built around FFmpeg, SoX, and Whisper with clear priorities (signal preservation is critical, tooling is low-medium). The forensic authentication rules like ENF analysis for timestamp verification show this goes beyond basic cleanup into actual investigative work.
npx skills add https://github.com/pproenca/dot-skills --skill audio-voice-recovery