When using the gap shortening or word gap editing tools, cuts are applied automatically based on duration thresholds. However, the resulting edit boundaries don't always align perfectly with the natural audio waveform — there are cases where a cut lands mid-breath, mid-consonant, or at a non-zero crossing, causing audible pops or unnatural transitions.

Request: Allow users (and the AI agent) to automatically fine-tune individual cut boundaries by dragging them automatically to align with waveform zero-crossings or natural silence points in the audio. Ideally this would include:

Waveform-snapping when dragging edit boundaries, so cuts snap to the nearest zero-crossing or silence

Per-cut adjustment in the timeline so individual gaps can be nudged frame-by-frame or sample-by-sample

Agent support so the AI can automatically detect and adjust cuts that don't land on clean audio boundaries

This would significantly improve audio quality for podcasts, interviews, and any heavily edited spoken-word content.