Content
57%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is well-structured and largely actionable with a complete example script, but suffers from redundancy between Key Features and Implementation Details, a missing validation/feedback loop for batch operations, and broken reference paths that undermine navigation to the otherwise well-organized bundle.
Suggestions
Fix the reference paths from `reference/` to `references/` and convert them to markdown links (e.g., [PREPROCESSING.md](references/PREPROCESSING.md)) so navigation resolves; also reference the unlinked ANALYSIS.md or remove it from the bundle.
Add explicit validation/verification checkpoints after batch operations — e.g., verify the Phy export folder and NWB file exist and are loadable after `si.export_to_phy` / `export_to_nwb`, with a fix-and-retry loop on failure — to lift workflow clarity above the batch-operation cap of 3.
Merge the "Key Features" bullets into "Implementation Details" (or vice versa) to remove the duplicated pipeline walkthrough and tighten conciseness, and add `neuropixels_analysis` (the `npa` wrapper) to the Dependencies section or clarify how it is provided so the example is fully runnable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient domain-specific guidance, but the "Key Features" section and the "Implementation Details" section walk the same pipeline in parallel, and there are a few explanatory asides ("Motion/drift can strongly degrade sorting quality") that could be trimmed; merging the two walkthroughs would tighten it, but it is not padded with concepts Claude already knows (which would warrant a 2). | 3 / 5 |
Actionability | Provides a largely executable, copy-paste-oriented script with real SpikeInterface calls and specific function/parameter names, but the example imports `neuropixels_analysis as npa` — a wrapper not listed in the Dependencies section and not established as installable — leaving a minor gap that keeps it below a 5. | 4 / 5 |
Workflow Clarity | Steps are clearly numbered and sequenced (1–8 in the example, 1–7 in Implementation Details) with one conditional checkpoint (drift > 10 µm), but the batch operations (sorting, curation labeling, Phy/NWB export) lack validation/verification or fix-and-retry feedback loops, so per the batch-operation cap workflow clarity cannot exceed 3. | 3 / 5 |
Progressive Disclosure | Good overall structure (overview SKILL.md plus 10 reference files, 6 scripts, 1 asset) with a dedicated reference list, but the body cites `reference/` (singular) while the bundle directory is `references/` (plural) so the paths do not resolve, references are backtick spans rather than navigable links, and `ANALYSIS.md` exists in the bundle but is never referenced — references are present but not correctly signaled. | 3 / 5 |
Total | 13 / 20 Passed |