Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when auditing a blog, news site, or YMYL content site for trust signals, drafting editorial policy page content, or evaluating whether a site meets Google's quality rater trustworthiness criteria.
.claude/skills/thedaviddias-editorial-policy/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -35% | 0% |
Google's E-E-A-T framework rates sites on Expertise, Authoritativeness, and Trustworthiness. Quality raters look for editorial standards pages as evidence of a trustworthy publication. Sites without visible editorial policies — especially YMYL sites — receive lower trust ratings, which suppresses search rankings.
Check if the site has a publicly accessible editorial policy page (commonly at /editorial-policy, /about/editorial-standards, or linked from the About page). Verify it covers: (1) how topics are chosen, (2) author qualifications, (3) fact-checking or review process, (4) how corrections are handled, and (5) conflicts of interest or funding disclosures.
Google's Search Quality Rater Guidelines specifically look for editorial standards as part of evaluating a site's Trustworthiness. For YMYL topics especially, the absence of an editorial policy is a clear signal to quality raters that the site cannot be trusted to publish accurate, expert-reviewed content, leading to lower quality ratings and reduced search visibility.
Check the site's footer links, About page, and individual article templates for a link to an editorial policy page. Verify the editorial policy URL returns a 200 status code and that the page contains substantive content — not just a boilerplate paragraph — covering research methodology, author qualifications, and correction policies.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/seo/editorial-policy
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-06 | fail→pass | 16,096 | 10,155 | -37% | 1 | 1 | 0% | 2,585 | 2,568 | -1% | 0 | 0 | — |
case-01 | fail→pass | 12,605 | 11,250 | -11% | 1 | 1 | 0% | 2,203 | 2,674 | +21% | 0 | 0 | — |
case-02 | pass→pass | 10,646 | 10,043 | -6% | 1 | 1 | 0% | 1,900 | 2,430 | +28% | 0 | 0 | — |
case-03 | pass→pass | 12,465 | 9,278 | -26% | 1 | 1 | 0% | 2,225 | 2,246 | +1% | 0 | 0 | — |
case-04 | pass→pass | 13,831 | 8,418 | -39% | 1 | 1 | 0% | 2,398 | 2,013 | -16% | 0 | 0 | — |
case-05 | pass→pass | 11,088 | 9,298 | -16% | 1 | 1 | 0% | 1,852 | 2,096 | +13% | 0 | 0 | — |
case-07 | pass→pass | 9,209 | 5,171 | -44% | 1 | 1 | 0% | 1,689 | 1,570 | -7% | 0 | 0 | — |
case-08 | fail→fail | 15,214 | 10,761 | -29% | 1 | 1 | 0% | 2,532 | 2,433 | -4% | 0 | 0 | — |
case-09 | pass→pass | 10,048 | 7,836 | -22% | 1 | 1 | 0% | 1,668 | 2,005 | +20% | 0 | 0 | — |
case-10 | pass→pass | 8,533 | 4,199 | -51% | 1 | 1 | 0% | 1,530 | 1,241 | -19% | 0 | 0 | — |
case-11 | pass→pass | 6,813 | 3,551 | -48% | 1 | 1 | 0% | 1,154 | 1,199 | +4% | 0 | 0 | — |
case-12 | pass→pass | 9,057 | 6,552 | -28% | 1 | 1 | 0% | 1,511 | 1,713 | +13% | 0 | 0 | — |
case-13 | fail→pass | 11,902 | 9,093 | -24% | 1 | 1 | 0% | 1,907 | 2,082 | +9% | 0 | 0 | — |
case-14 | fail→pass | 15,221 | 9,046 | -41% | 1 | 1 | 0% | 2,422 | 2,109 | -13% | 0 | 0 | — |
case-15 | fail→pass | 8,729 | 2,252 | -74% | 1 | 1 | 0% | 1,467 | 952 | -35% | 0 | 0 | — |
case-16 | pass→pass | 10,030 | 11,307 | +13% | 1 | 1 | 0% | 1,816 | 2,472 | +36% | 0 | 0 | — |
case-17 | pass→pass | 11,680 | 5,967 | -49% | 1 | 1 | 0% | 1,896 | 1,718 | -9% | 0 | 0 | — |
case-18 | pass→pass | 10,174 | 7,610 | -25% | 1 | 1 | 0% | 1,742 | 1,872 | +7% | 0 | 0 | — |
case-19 | pass→pass | 11,822 | 9,843 | -17% | 1 | 1 | 0% | 2,081 | 2,205 | +6% | 0 | 0 | — |
case-20 | pass→fail | 14,467 | 14,097 | -3% | 1 | 1 | 0% | 2,651 | 3,225 | +22% | 0 | 0 | — |
case-21 | pass→pass | 11,310 | 11,442 | +1% | 1 | 1 | 0% | 2,312 | 2,931 | +27% | 0 | 0 | — |
case-22 | pass→pass | 12,680 | 14,359 | +13% | 1 | 1 | 0% | 2,557 | 3,427 | +34% | 0 | 0 | — |
case-23 | pass→pass | 15,827 | 10,678 | -33% | 1 | 1 | 0% | 3,007 | 2,741 | -9% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.