Content
82%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 dense, factual, well-sectioned reference that is almost entirely Netlify-specific knowledge Claude cannot infer, with copy-paste-ready code for every common case and genuine footguns surfaced up front. The main improvement levers are trimming the meta-commentary (ctx-gen preamble, HTML gap comments) and moving the exhaustive @netlify/cache API detail to a reference file.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Phrases like "Cache API silently drops responses over 10 MB", "Deletes are eventually consistent", and the exact Cache-Status values are non-obvious, high-value facts with essentially no explanation of concepts Claude already knows — but trimmable tokens remain: the ctx-gen house-rules preamble ("merged into the rendered skill by ctx-gen and never generated"), the trailing HTML gap comments, and a second full vary illustration inside the cacheHeaders example. These are minor instances of over-explanation/padding rather than the 'every token earns its place' ideal, so 4 not 5; it is well above the 'mostly efficient but includes some unnecessary explanation' anchor 3. | 4 / 5 |
Actionability | Nearly every section gives copy-paste-ready code or commands — the full `Netlify-CDN-Cache-Control` function example, three complete `purgeCache` variants including the token-based Lambda form, an executable curl for the purge API, the `caches.open/match/put` handler with `config`, and `cacheHeaders`/`fetchWithCache` snippets — plus exact limits ("100 lookups, 20 insertions/deletions", "twice per 5s", "≤1024 chars/tag"). Only two contextual fragments (the `needsRevalidation` block, the `getCacheStatus` one-liner) lack imports/context, which keeps it at anchor 5's 'fully executable, covers common cases' rather than dropping to anchor 4's 'minor gaps' since the fragments are advanced edge cases, not the common path. | 5 / 5 |
Workflow Clarity | The skill is reference-style rather than a multi-step procedure, and it sequences its single task well: 'read footguns first' → pick the header ('Reach for this') → set directives → verify on a deployed URL via Cache-Status ('repeat requests to warm one' gives an explicit feedback loop for the empty-cache case). Validation for the destructive purge operation is only implicit (the 202 status in examples) rather than an explicit 'verify the purge succeeded' checkpoint, which is the minor gap that keeps this at anchor 4 ('most checkpoints present') instead of 5; it is clearly above anchor 3's missing-checkpoints level. | 4 / 5 |
Progressive Disclosure | No bundle files exist (no references/, scripts/, or assets/), so everything is inline in a well-organized single file: clearly signaled sections (Footguns, Directives, Netlify-Vary, purge, Cache API, @netlify/cache, Durable, Debugging) with the most-urgent content first and an external pointer only where appropriate (docs.netlify.com for the purge API identifier mapping). At ~330 lines, the full `@netlify/cache` API details could arguably live in a reference file, which is the 'minor organization gap' of anchor 4 — better than anchor 3's 'content that should be separate is inline' since the structure is tight and navigable, but not the clean overview-plus-references split of anchor 5. | 4 / 5 |
Total | 17 / 20 Passed |