Runs spike sorting pipelines using the SpikeLab library (Kilosort2, Kilosort4, RT-Sort). Handles configuring and executing sorting jobs, curating units, inspecting and visualizing results. For stimulation experiments, runs artifact removal via preprocess_stim_artifacts / sort_stim_recording and then the appropriate sorter. Use when the user wants to sort recordings, curate units, or analyze sorting outputs.
Implements spike train data analysis and visualization using the SpikeLab library. Handles writing analysis and visualization scripts, managing results, and updating repo maps. Use when you want to load, analyze, or visualize spike train data with SpikeLab.
Integrates analysis code into the SpikeLab library. Takes user-provided scripts, identifies which computations already exist in the library, rewrites the script to use existing methods, integrates novel computations as new library methods, writes tests, exposes through MCP, and submits a PR. Use when the user wants to promote analysis code into the library.
Answers questions about SpikeLab analyses, neuroscience concepts, and how to interpret results. Read-only — never writes or executes code. Use when a user asks what an analysis does, how it works, what a result means, or wants to learn about the underlying neuroscience.
Generates and updates the repo map files that document the SpikeLab library API. Run after installing or updating the library. Not intended for use during analysis sessions.