seo-freshness
Audit and repair freshness & temporal signals on a page — reconcile visible publish/update dates with schema datePublished/dateModified, flag staleness against topic volatility, and inject honest dateModified. Module M13. Feeds both the Search SEO and AI Visibility scores.
适合你,如果负责网站SEO,需要确保页面日期信号准确且符合时效性要求
用别的 agent?下载 .zip 解压,把文件夹放进它的技能目录
~/.claude/skills/(项目级 .claude/skills/)~/.codex/skills/npx oh-my-skill add hainrixz/claude-seo-ai/seo-freshnesscurl -fsSL https://oh-my-skill.com/install.sh | bash -s -- hainrixz/claude-seo-ai/seo-freshnessnpx oh-my-skill verify hainrixz/claude-seo-ai/seo-freshness怎么用
技能原文 SKILL.md
seo-freshness (M13)
Freshness is a recency signal both classic ranking systems (Query Deserves Freshness) and AI answer engines weigh — Perplexity in particular favours recently-updated sources when citing. Date fields tie directly to Article schema (cross-check M5); see references/schema-tier1.md for the date rules.
Audits
Working from the PageSnapshot (rendered_dom if present, else raw_html):
- Visible dates: detect on-page "Published" / "Updated on" / "Last reviewed" patterns and their values (ISO or human-readable).
- Schema dates: parse
datePublished/dateModifiedfrom JSON-LDArticle/BlogPosting/NewsArticle. - Agreement: visible date and schema date must match; flag mismatches and schema dates with no visible counterpart (AI engines distrust hidden-only dates).
- Staleness: estimate content age (most recent reliable date) vs topic volatility — fast-moving topics (prices, tooling, "best X 2026", regulations) decay faster than evergreen reference content. Report stale, not just old.
- Pattern hygiene: "updated on" with no substantive content change is a freshness anti-pattern — note it, never recommend it.
Fixes
- AUTO (
fixable: auto): inject a missingdateModifiedinto existing Article schema as an additive diff forfix. Never backdate to a false date — use the verifiable last-change date (e.g. Last-Modified header / repo mtime / today) or leave a clearly-markedTODOplaceholder the user confirms. - PROPOSED (
fixable: proposed): surface visible-vs-schema date mismatches with the corrected value as a draft requiring per-item accept; never auto-rewrite a date the user must verify. - ADVISORY (
fixable: advisory): recommend a genuine content refresh for stale-on-volatile pages — the tool never writes editorial content. Never fabricate dates or invent an update that did not happen.
Verification
dom_assert: visible date string present and parses; matches schema value.schema_validator:datePublished/dateModifiedpresent, valid ISO 8601,dateModified >= datePublished.header_check: HTTPLast-Modifiedheader corroborates the claimed modification date.- When the live tier (header fetch / validator) is unavailable, status is
needs_api, never a falsepass.
Findings
Findings conform to schema/finding.schema.json. Examples:
M13.datemodified.missing— Article schema withdatePublishedbut nodateModified(statusfail, severity 3,fixable: auto, axisboth, confidencedirectional).M13.dates.visible_schema_mismatch— visible "Updated May 2025" vs schemadateModified: 2023-01-10(statuswarn, severity 3,fixable: proposed, axisboth, confidencedirectional).M13.content.stale_volatile— "best X 2024" page unchanged for 2 years on a fast-moving topic (statuswarn, severity 2,fixable: advisory, axisboth, confidencedirectional).
Each finding: evidence.observed quotes the page (date string + selector); verification.reproduce is a runnable assertion (e.g. node scripts/check-freshness.mjs --url <u>); expected_impact is banded + confidence-tagged, with any published number confined to rationale with a citation.
Honesty
- Freshness is a contextual signal, not a universal ranking boost — refreshing evergreen content rarely moves rankings, and a
dateModifiedbump without a real content change is detectable and adds no durable value. Confidence here is directional, neverestablished; never present a date edit as a guaranteed ranking gain or backdate to fake recency.