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.
A high-quality, dense reference: every section delivers non-obvious, Netlify-specific facts with executable code, exact limits, and a strong diagnostic checklist. The main defects are the duplicated trailing 'house rules' section (which re-states seven rules already in the body) and the absence of explicit post-setup verification checkpoints.
Suggestions
Deduplicate the trailing 'Netlify house rules (forms)' section: its seven rules (SSR POST target, Link-header paging, skeleton file, Astro on-demand routes, Akismet false positive, extensionless action paths, documented-surfaces-only) already appear in the body sections — keep one copy and drop the other to reclaim ~30 lines of tokens.
Add an explicit verification checkpoint to the setup workflows, e.g. after deploying a new form: 'submit a real test (real email, full sentences — test submissions get spam-flagged) and confirm it appears in Forms > Verified/Spam before finishing', turning the reactive debugging section into a proactive feedback loop.
Remove or externalize build artifacts that consume context without aiding the model: the ctx-gen merge comment ('edit system.md, not this section') and the two trailing 'Gap:' HTML comments could move to the human-owned source file instead of the rendered skill.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and assumes Claude's competence (no basic-concept explanations), but the trailing "Netlify house rules (forms)" section restates ~7 rules already covered verbatim above (SSR POST target, Link-header paging, skeleton file, Astro on-demand routes, Akismet false positive, extensionless action paths), and the ctx-gen merge preamble plus gap comments add redundant tokens. This fits 'Mostly efficient but... could be tightened'; it is not score 4 because the duplication is a substantial trim target (~17% of the body), not a minor instance of over-explanation. | 3 / 5 |
Actionability | Guidance is fully executable throughout: copy-paste HTML/JSX/JS code blocks, exact attributes (data-netlify, netlify-honeypot, data-netlify-recaptcha), exact env vars (SITE_RECAPTCHA_KEY/SECRET), exact limits (8 MB, 30 s, 24 h), exact UI paths, and exact API query (?state=spam). This matches 'Fully executable; copy-paste ready code or commands; specific examples cover the common cases'. | 5 / 5 |
Workflow Clarity | Scenario workflows are clearly sequenced (e.g., JS-rendered forms: skeleton file + matching form-name input, with SSR and Astro gotchas called out) and the debugging section is an explicit error-recovery checklist with feedback loops (check Spam list and mark verified; confirm detection is enabled and redeploy; verify skeleton field names). It falls short of 5 because the setup workflows lack an explicit verification checkpoint (e.g., 'submit a test and confirm it appears after deploy') — checkpoints appear only reactively in the debugging section. | 4 / 5 |
Progressive Disclosure | The body is well organized into task-oriented sections with clear headers, flagged warnings, and clearly labeled external links, and each section is short enough that inlining is appropriate for a single-file skill with no bundle files. This fits 'Good structure; most content is appropriately placed... minor organization gaps'; it is not score 5 because nothing is layered into reference files and the duplicated house-rules appendix is an organizational defect, and not score 3 because the content is not a reference dump that clearly belongs in separate files. | 4 / 5 |
Total | 16 / 20 Passed |