v0.1.0 — look, speak, send

First public release. Gaze + voice as a multimodal input layer for macOS — look at something on screen, speak, and your gaze-resolved intent goes to Claude. Swift, on-device eye tracking.

What's Changed

• Phase 1: webcam probe (Vision landmarks + gaze feature) by @Tatendaz in #1
• Phase 2: calibration & on-screen gaze cursor by @Tatendaz in #2
• Phase 3: fixation event stream by @Tatendaz in #3
• Recalibrate-on-drift (head-movement robustness) by @Tatendaz in #4
• Phase 4: voice fusion — utterance event stream by @Tatendaz in #5
• Phase 5: element resolution — own-canvas named gaze targets by @Tatendaz in #7

New Contributors

• @Tatendaz made their first contribution in #1

Full Changelog: https://github.com/Tatendaz/Vergance/commits/v0.1.0