Content
75%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 an exceptionally dense, actionable reference: executable code for every task, exact UI navigation paths, and genuinely field-learned gotchas (SSR POST interception, extensionless success pages, Akismet false positives) that Claude would not know. Its main weaknesses are the verbatim-duplicated 'house rules' section and meta comments that spend tokens without helping the executing agent.
Suggestions
Remove or merge the 'Netlify house rules' section — its seven points restate the SSR POST-target, skeleton-file, Astro, Akismet, Link-header, and extensionless-path guidance already covered verbatim in the main sections, roughly doubling that content.
Delete the 'Gap:' HTML comments (or move them to a maintainer note outside SKILL.md) — they document known coverage gaps for maintainers but consume context tokens for the executing agent.
Consider moving rarely-needed detail such as the submission-summary field-type heuristics and the upload limits table into a reference file so the main body leads with the highest-frequency tasks.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly dense and lean, but the 'Netlify house rules' section (points 1–7) restates the SSR POST-target, skeleton-file, Astro, Akismet, Link-header, and extensionless-path guidance almost verbatim from earlier sections, and the 'Gap:' HTML comments add tokens without aiding the executing agent — so it fits anchor 3 ('could be tightened') rather than anchor 2, since the rest carries no padded explanation of concepts Claude already knows. | 3 / 5 |
Actionability | Every task has copy-paste-ready code — static HTML form, `public/__forms.html` skeleton, React form, AJAX fetch with URLSearchParams, no-header file upload, honeypot markup, subject-line input — plus exact UI paths (Forms > Submission notifications > Add notification) and concrete limits. This is fully executable guidance covering the common cases, matching anchor 5. | 5 / 5 |
Workflow Clarity | Multi-step tasks are clearly sequenced (the two required pieces for framework forms, numbered with explicit warnings), and the debugging section provides real feedback loops (check the Spam list → mark verified; notification validation error → invite the team member; verify the skeleton file exists with exactly-matching names). It is anchor 4 rather than 5 because there is no single coherent top-level workflow with explicit validate-before-proceed checkpoints — the skill is a task reference rather than one validated pipeline. | 4 / 5 |
Progressive Disclosure | With no bundle files present, this is a single-file skill with well-organized section headers and clearly labeled external documentation links (Next.js v5 breaking changes, OpenAPI reference, support guide). Structure is good with minor organization gaps — the duplicated house-rules block and inline gap comments keep it from anchor 5's 'appropriately split' ideal. | 4 / 5 |
Total | 16 / 20 Passed |