seo-content-audit — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited seo-content-audit (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
A page can pass every structural check, load fast, validate every schema, and still rank nowhere — because the content doesn't match search intent, the headings bury the answer, the byline says "Admin," or the top 10 already covers everything this page says. This skill diagnoses the on-page layer that the structural audit can't see: how the content is shaped against the SERP it wants to win.
The audit cross-references the target page against the actual SERP — top 10 results, "People Also Ask"-style questions, AI Overview presence where it can be observed — and returns findings where the page falls short of what's already winning. Information gain (what does this page add that the top 10 doesn't) is the priority test: Google's May 2026 AI optimization guide asks for non-commodity, first-hand content, and third-party readings of the March and May 2026 core updates saw original sources gain over aggregators.
On-page content audit is intent-versus-execution:
The audit is not "compare to a checklist." It's "compare to the SERP." Generic on-page advice (keyword in title, keyword in H1) is necessary but not sufficient in 2026. Sufficient is matching the format, depth, and experience signal of the pages that already rank.
knowledge(command="read", source="<url>") — title, meta, headings, body, byline, JSON-LD.knowledge — for direct comparison.Run all four in one block. Don't serialize. The web search tool is not Google Search: AI Overview / AI Mode presence, featured-snippet holders, and People Also Ask are Unverified unless the user supplies a screenshot or Search Console data.
| # | Check | How to verify | Severity if failed |
|---|---|---|---|
| 1 | Intent match | Read the target keyword aloud — what does the searcher want? Compare to page intent. | Critical if format mismatches dominant SERP intent |
| 2 | Title tag | 50–60 chars, primary keyword front-loaded, compelling (clear value, not generic) | Moderate per failed dimension |
| 3 | Meta description | 150–160 chars, keyword present, answers "why click this result" | Moderate per failed dimension |
| 4 | H1 | Present, unique, semantically related to title (not identical), describes the page's promise | Critical if missing or duplicate |
| 5 | H2 structure | Headings describe what each section answers; question phrasing where it reads naturally | Minor if headings are vague labels that hide the content |
| 6 | Answer placement | The direct answer to the section's question appears early, before supporting detail | Moderate per key section that buries its answer |
| 7 | Self-contained sections | Each major point reads on its own without reconstructing it from several paragraphs | Minor; don't recommend splitting content into fragments — Google says chunking isn't needed |
| 8 | Data density | At least one concrete number, percentage, or statistic per major section | Minor per data-free section; Moderate if entire piece lacks specifics |
| 9 | FAQ section | Present when users genuinely ask follow-up questions (informational and commercial intent) | Minor if absent and intent is suitable; FAQ rich results ended May 7, 2026, so judge it on reader value only |
| 10 | Author byline | Real name + role + credentials + outbound link (LinkedIn / publication / org bio). "Admin" or anonymous = Critical | Critical if anonymous or generic-name attribution |
| 11 | Experience proof | Original screenshots / dashboards / photos from the field / named tools+versions / time-boxed case-study language | Critical if pure explainer with no experience signal on YMYL or expertise-required topic |
| 12 | Information gain vs top 10 | Identify the specific thing this page adds (original data, contrarian take, novel synthesis, practitioner detail) that the top 10 doesn't | Critical if the page is reconstructable from the top 10 (information-gain-zero) |
| 13 | Internal linking | 5–10 contextual links per piece; descriptive anchors (not exact-match); pillar/cluster pattern present where applicable | Moderate if under 3 internal links or all anchor text is generic |
| 14 | Schema readiness for content | Article schema present with author/datePublished/dateModified; markup matches visible text | Moderate per invalid or contradicting markup; Google states AI features need no special schema |
| 15 | Primary-source posture | Page presents primary data/experience/original synthesis OR cites primary sources where it relies on them | Critical if the page is an intermediary roundup of other rankers without added signal |
Information gain is the highest-leverage content check — when in doubt, weight it Critical.
Before scoring, fetch and skim 3–5 top-ranking pages. Identify everything the SERP already says. Then score the target on:
If you cannot name what the target page adds that the top 10 doesn't, that is the finding: "Information-gain-zero — page reconstructable from current top 10. Highest-priority fix."
Google's May 2026 AI optimization guide says AI Overviews and AI Mode need no special phrasing, chunking, llms.txt, or schema, and that a page per query variation counts as scaled content abuse. Audit structure for readers, not for a machine:
Never infer information gain from one source; never grade an anonymous byline before checking if there's an author bio page linked.
Use the canonical output structure from the agent that runs this skill. Every finding includes: severity, evidence (URL section, line, the missing element, the SERP comparison point), why it matters, fix direction. Save as seo-audit-[slug]-[YYYY-MM-DD].md in working directory.
Within-domain pairings:
External authoritative sources:
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.