Content
81%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 a strong, executable audit playbook: framework-aware detection commands, severity-weighted scoring, an evidence rule, and a verification checklist make it highly actionable with a clear workflow. Weaknesses are minor — the anti-rationalization table is trimmable padding, and the $TARGET/$ARGUMENTS placeholder inconsistency should be cleaned up.
Suggestions
Compress the Anti-Rationalization table to one or two lines (or drop it) — Claude does not need to be pre-argued into SEO's importance, and it is the largest block of non-operational text.
Define the target variable once (e.g. "TARGET=$ARGUMENTS") and use it consistently, replacing the bare "." in the Step 5 robots.txt/sitemap finds so every command is verbatim runnable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and operational — check tables, severity ratings, and bounded grep commands with head limits earn their tokens. Not 5 because the 7-row "Anti-Rationalization" table argues for SEO's importance (knowledge Claude already has) and could be trimmed to one line; not 3 because there is little concept explanation and almost no padding elsewhere. | 4 / 5 |
Actionability | Every step ships copy-paste-ready bash (grep/find with --include filters, head caps) plus an explicit scoring formula ("CRITICAL finding: -25 points..."), fitting "mostly executable guidance". Not 5 because $TARGET is used throughout but never defined (the declared target is $ARGUMENTS), and one command uses "." instead of the target — minor gaps a reader must patch. | 4 / 5 |
Workflow Clarity | Six clearly sequenced steps (scope → technical → content → CWV → crawlability → scoring) with an explicit "Verification Protocol" checklist ("Before claiming the audit is complete: 1-6"), including per-finding evidence requirements and a page-count floor. This matches the anchor with explicit validation steps and a checklist for a complex process; the read-only nature of the audit means no error-recovery loop is needed. | 5 / 5 |
Progressive Disclosure | No bundle files exist and the single-file body is well-sectioned (numbered steps, tables, an output template), so navigation is easy and nothing is buried. Not 5 because at ~230 lines the per-category check catalogs could be split into one-level-deep reference files, and the under-50-line exception does not apply; not 3 because the structure is good and no content is misplaced or inlined that clearly demands separation. | 4 / 5 |
Total | 17 / 20 Passed |