Content
36%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.
A well-sectioned but bloated pattern reference: the structural decisions (split settings, custom QuerySet, service layer, N+1 prevention) are genuinely useful, yet they are buried in boilerplate Claude already knows and non-executable snippets with undefined imports and phantom fields. No validation guidance accompanies the destructive/bulk operations, and the entire multi-domain catalog lives in one file instead of being progressively disclosed.
Suggestions
Trim standard Django boilerplate (INSTALLED_APPS/MIDDLEWARE lists, base AbstractUser fields, stock serializer scaffolding) to the pattern decisions that Claude would not infer on its own.
Make code examples executable: import or define 'env', 'slugify', 'logger', 'timezone', 'Count', 'Cart', and 'PaymentGateway', and reconcile referenced fields ('is_featured', 'created_by') with the shown models.
Add validation/feedback-loop guidance around bulk operations (e.g. verify counts or wrap bulk_delete in a transaction with a dry-run check) and split the DRF, settings, and performance sections into one-level-deep reference files with the SKILL.md kept as a concise overview.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~730-line body inlines substantial boilerplate Claude already knows — full INSTALLED_APPS/MIDDLEWARE lists, a stock AbstractUser model, standard ModelSerializer patterns — matching anchor 2 ('several unnecessary explanations or padded sections'). Not a 1 because there is no prose conceptual explanation and the pattern choices themselves (split settings, service layer, custom QuerySet) do add value. | 2 / 5 |
Actionability | Code examples are concrete and idiomatic but not executable as written: 'env()', 'slugify', 'logger', 'timezone', 'Count', 'Cart', and 'PaymentGateway' are never imported or defined, model fields are elided ('# ... fields ...'), and snippets reference fields that don't exist on the shown models ('is_featured', 'created_by'). This matches anchor 3 ('concrete guidance but incomplete; missing key details'), well short of copy-paste-ready anchor 4. | 3 / 5 |
Workflow Clarity | There is only an implicit ordering (structure → models → serializers → views → services) with no explicit steps or checkpoints, and the bulk operations section shows 'Product.objects.filter(stock=0).delete()' and bulk_create/bulk_update with no validation or feedback loop — anchor 2 ('rough sequence present but many gaps; validation absent'), and the destructive/batch cap at 3 is not approached anyway. | 2 / 5 |
Progressive Disclosure | Sections are well-organized with headers and a Quick Reference table, but everything lives inline in one monolithic ~730-line file — settings examples, full model/serializer/ViewSet code, caching, signals, middleware, and performance sections that clearly belong in separate reference files, and no bundle references exist at all. This sits between anchor 2 ('content that clearly belongs in separate files is inlined') and anchor 3 ('some structure but content that should be separate is inline'), so 3 is the best fit given the genuinely good in-file structure. | 3 / 5 |
Total | 10 / 20 Passed |