neuroskill-search

What problem does it solve?

Mapping current EEG states to similar historical moments and comparing two sessions is tedious because all metrics and embeddings live across scattered history, so this skill surfaces relevant neighbors and delta trends automatically.

Core Features & Use Cases

• Neural similarity search: The search command runs an ANN query over five-second embeddings, highlights temporal distributions and metrics, and lets you auto-range to the last session or specify exact start/end times.
• Session comparison: The compare command averages metrics for two ranges, reports deltas with trend directions, flags improvements and declines, and enqueues a 3D UMAP projection for spatial state comparison.
• Cross-modal follow-ups: After retrieving neighbors, bridge to screenshots, label search, or the interactive graph tool to contextualize similar focus states from other sensors.
• Use Case: To track when you last reached a relaxed focus state, search your entire history for matching embeddings and then compare that session to today to see which metrics improved.

Quick Start

Ask the skill to search for the five closest EEG neighbors from your most recent block to see similar focus states.