Conciseness

The body is dense with pipeline-specific executable code and never explains concepts Claude already knows (no 'what is spike sorting' padding). However, 'Overview' and 'When to Use This Skill' largely restate the frontmatter description, and 'Bundled Resources' plus 'Detailed Reference Guides' list the same reference files twice — minor trimming that keeps it at anchor 4 rather than 5.

Actionability

Nearly everything is copy-paste-ready: 'si.read_spikeglx(\'/path/to/data\', stream_id=\'imec0.ap\')', the full preprocessing chain with parameters, 'si.run_sorter(\'kilosort4\', rec, folder=\'ks4_output\')' plus CPU alternatives, all 'analyzer.compute(...)' extensions with arguments, a thresholded curation query, and CLI invocations for every bundled script. Specific examples cover the common cases end-to-end, matching anchor 5.

Workflow Clarity

The 8-step 'Standard Analysis Workflow' is clearly sequenced and includes an explicit checkpoint — step 2 says 'Always check drift before spike sorting' with a '> 10' micron threshold gating whether correction is applied. It stops short of anchor 5 because later stages lack verification (no check that sorting output exists, no validation that the NWB export loads) and there are no feedback loops for error recovery.

Progressive Disclosure

All 10 referenced reference files exist and reference only external URLs (one level deep), scripts and the asset template are documented with CLI usage, and the SKILL.md works as an overview with quick-start inline. Minor organization gaps hold it at anchor 4: the reference files are listed twice (Bundled Resources section and Detailed Reference Guides table), and bundled scripts 'explore_recording.py' and 'neuropixels_pipeline.py' are never mentioned.

Specificity

The description lists multiple concrete, tool-named actions spanning the full pipeline — 'Load SpikeGLX/OpenEphys data, preprocess, motion correction, Kilosort4 spike sorting, quality metrics, Allen/IBL curation, AI-assisted visual analysis' — which matches the anchor-5 example's comprehensive multi-action coverage. It is not a 4 because coverage runs end-to-end (ingest through curation) with specific named tools rather than leaving minor gaps.

Completeness

It explicitly answers both questions with concrete trigger phrases: the 'what' is the action list ('Load SpikeGLX/OpenEphys data... AI-assisted visual analysis') and the 'when' is the clause 'Use when working with neural recordings... or when the user mentions Neuropixels...'. This mirrors the anchor-5 example structure almost exactly, so a 4 ('when could be more explicit') does not fit.

Trigger Term Quality

The 'when' clause covers good natural keywords: 'neural recordings, spike sorting, extracellular electrophysiology, Neuropixels, SpikeGLX, Open Ephys, Kilosort, quality metrics, or unit curation'. It falls short of anchor 5 because common variations users actually say are missing — 'ephys', file extensions (.ap.bin, .meta, .nwb), and 'drift'/'spike sorted data' — while exceeding anchor 3's 'missing common variations' by including both tool names and workflow terms.

Distinctiveness Conflict Risk

Highly specific niche triggers — 'Neuropixels', 'SpikeGLX', 'Open Ephys', 'Kilosort', 'unit curation' — are unlikely to appear when another skill is needed, giving a clear niche with minimal conflict risk per anchor 5. It is written in third person, so no voice penalty applies.