The Agent Skills Directory

[COMMAND_EXECUTION]: The skill leverages python-based command execution via Bash(python:*) and Bash(uv:*) to perform complex neural data processing workflows including filtering, motion correction, and spike sorting using the SpikeInterface framework.
[EXTERNAL_DOWNLOADS]: The documentation provides instructions for installing established scientific software and libraries (e.g., SpikeInterface, Kilosort, NeuroConv) from standard registries like PyPI to support the analysis pipeline. These references target well-known research tools and are documented for user setup.
[INDIRECT_PROMPT_INJECTION]: The skill defines a data ingestion surface by loading Neuropixels recordings and metadata files (.bin, .meta, .nwb). The ingestion points are located in scripts/neuropixels_pipeline.py and scripts/preprocess_recording.py. While this presents a theoretical attack surface, the specific nature of the data (electrophysiological voltage traces) and the scientific context minimize potential for malicious instruction injection.
[SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration were detected. The skill follows security best practices by providing explicit warnings about illustrative pseudocode in references/api_reference.md and utilizing modern environment management tools.

alterlab-neuropixels