@rankcli/agent-runtime 0.0.38 → 0.0.39

package/dist/index.mjs

@@ -2824,7 +2824,7 @@ var ISSUE_DEFINITIONS = {
     severity: "warning",
     category: "performance",
     title: "HTML document too large",
-    description: "HTML size exceeds 100KB.",
+    description: "HTML size exceeds 300KB.",
     impact: "Increases load time and memory usage.",
     howToFix: "Reduce inline CSS/JS, remove unnecessary HTML."
   },
@@ -5559,7 +5559,7 @@ async function analyzePerformance(url) {
         details: { ttfb: `${ttfbTime}ms` }
       });
     }
-    if (htmlSize > 100 * 1024) {
+    if (htmlSize > 300 * 1024) {
       issues.push({
         ...ISSUE_DEFINITIONS.HTML_TOO_LARGE,
         affectedUrls: [url],
@@ -7717,7 +7717,7 @@ function analyzeLocalSEO(html, url) {
   const detectedPhones = detectPhones(html);
   const detectedAddresses = detectAddresses(html);
   const socialProfiles = detectSocialProfiles(html);
-  if (!hasLocalBusiness) {
+  if (!hasLocalBusiness && (detectedPhones.length > 0 || detectedAddresses.length > 0)) {
     issues.push({
       ...ISSUE_DEFINITIONS.LOCAL_BUSINESS_SCHEMA_MISSING,
       affectedUrls: [url]
@@ -8778,7 +8778,17 @@ function analyzeKeywords(html, url) {
   const sortPhrases = (phrases, n) => {
     return Array.from(phrases).filter(([_, count]) => count >= 2).sort((a, b) => b[1] - a[1]).slice(0, 10).map(([phrase, count]) => ({ phrase, count, wordCount: n }));
   };
-  const potentialStuffing = topKeywords.filter((k) => k.density > 3).map((k) => k.word);
+  const brandLabels = /* @__PURE__ */ new Set();
+  try {
+    const host = new URL(url).hostname.toLowerCase().replace(/^www\./, "");
+    for (const label of host.split(".").slice(0, -1)) brandLabels.add(label);
+  } catch {
+  }
+  for (const name of [$('meta[property="og:site_name"]').attr("content"), $('meta[name="application-name"]').attr("content")]) {
+    for (const token of tokenize(name || "")) brandLabels.add(token);
+  }
+  const isBrandWord = (word) => [...brandLabels].some((label) => label.includes(word) || word.includes(label));
+  const potentialStuffing = totalWords >= 200 ? topKeywords.filter((k) => k.density > 4 && !isBrandWord(k.word)).map((k) => k.word) : [];
   if (potentialStuffing.length > 0) {
     issues.push({
       ...ISSUE_DEFINITIONS.KEYWORD_STUFFING,
@@ -9004,13 +9014,6 @@ function checkAMP(html, url) {
       }
     };
   }
-  const isArticle = $("article").length > 0 || $('meta[property="og:type"][content="article"]').length > 0 || html.toLowerCase().includes("blog") || html.toLowerCase().includes("news");
-  if (isArticle) {
-    issues.push({
-      ...ISSUE_DEFINITIONS.AMP_MISSING,
-      affectedUrls: [url]
-    });
-  }
   return {
     issues,
     data: {
@@ -9719,13 +9722,13 @@ async function analyzeSiteMaturity(html, url) {
     contentToAuthorityRatio
   };
   const { strategy, recommendations } = determineStrategy(maturityLevel, authorityIndicators, contentMetrics);
-  if (maturityLevel === "new") {
+  if (maturityLevel === "new" && ageSource === "wayback" && estimatedAgeDays !== null) {
     issues.push({
       code: "NEW_SITE_STRATEGY",
       severity: "notice",
       category: "content",
       title: "New website detected - Authority-first strategy recommended",
-      description: "Your site appears to be new (<90 days). Standard SEO advice won't work well for new sites.",
+      description: `The Wayback Machine first saw this site ${estimatedAgeDays} days ago. Standard SEO advice won't work well for a site under 90 days old.`,
       impact: "Creating lots of content without authority is wasted effort. Content from new sites rarely ranks.",
       howToFix: "Focus on building 10+ quality backlinks to ONE great page before expanding content. See strategic recommendations.",
       affectedUrls: [url],
@@ -9736,24 +9739,7 @@ async function analyzeSiteMaturity(html, url) {
       }
     });
   }
-  if (contentToAuthorityRatio === "content-heavy") {
-    issues.push({
-      code: "CONTENT_AUTHORITY_IMBALANCE",
-      severity: "warning",
-      category: "content",
-      title: "Content-to-authority imbalance",
-      description: `Site has ~${indexedPages} pages but limited backlink signals. Content is unlikely to rank.`,
-      impact: "Most of your content may be invisible to search engines due to low domain authority.",
-      howToFix: "Stop publishing new content. Focus 100% on link building until you see ranking improvements.",
-      affectedUrls: [url],
-      details: {
-        indexedPages,
-        contentToAuthorityRatio,
-        maturityLevel
-      }
-    });
-  }
-  if (maturityLevel === "growing" && !htmlSignals.hasBlog) {
+  if (maturityLevel === "growing" && ageSource === "wayback" && !htmlSignals.hasBlog) {
     issues.push({
       code: "GROWING_SITE_NO_BLOG",
       severity: "notice",
@@ -10247,8 +10233,8 @@ function analyzeRendering(html) {
   const hasSemanticStructure = hasH1 && (hasMain || hasArticle);
   const title = $("title").text().trim();
   const h1Text = $("h1").first().text().trim();
-  const firstParagraph = $("p").first().text().trim();
-  const keyContentInHTML = title.length > 10 && h1Text.length > 5 && firstParagraph.length > 50;
+  const hasSubstantialParagraph = $("p").toArray().some((p) => $(p).text().trim().length > 50);
+  const keyContentInHTML = title.length > 0 && h1Text.length > 0 && hasSubstantialParagraph;
   const EXPECTED_MIN_CONTENT = 1e3;
   const htmlContentRatio = Math.min(100, mainContentCharCount / EXPECTED_MIN_CONTENT * 100);
   let aiReadableScore = 50;
@@ -10468,9 +10454,9 @@ function analyzeClientRendering(html, url) {
       howToFix: "Ensure the initial HTML contains: meaningful <title>, descriptive <h1>, and a substantial first paragraph with your key message.",
       affectedUrls: [url],
       details: {
-        hasTitle: $("title").text().trim().length > 10,
-        hasH1: $("h1").text().trim().length > 5,
-        hasFirstParagraph: $("p").first().text().trim().length > 50
+        hasTitle: $("title").text().trim().length > 0,
+        hasH1: $("h1").text().trim().length > 0,
+        hasSubstantialParagraph: $("p").toArray().some((p) => $(p).text().trim().length > 50)
       }
     });
   }
@@ -11179,7 +11165,9 @@ async function analyzeBingOptimization(html, url) {
     detectedIssues.push("No Bing site verification");
     issues.push({
       code: "BING_NOT_VERIFIED",
-      severity: "warning",
+      // A notice: verification done through Search Console import or a DNS record this run
+      // could not see leaves no trace in the HTML, so this cannot be a warning.
+      severity: "notice",
       category: "ai-readiness",
       title: verification.dnsChecked ? "Site not verified with Bing Webmaster Tools" : "Bing Webmaster Tools verification not visible in the HTML",
       description: verification.dnsChecked ? "No msvalidate.01 meta tag in the HTML, no MS= TXT record on the domain and no /BingSiteAuth.xml file. ChatGPT Search uses Bing's index. Verifying your site with Bing Webmaster Tools ensures proper indexing and access to Bing-specific SEO tools." : "No msvalidate.01 meta tag in the HTML and no /BingSiteAuth.xml file. The site may be verified through a DNS record instead, which this run could not check. ChatGPT Search uses Bing's index. Verifying your site with Bing Webmaster Tools ensures proper indexing and access to Bing-specific SEO tools.",
@@ -11230,7 +11218,7 @@ async function analyzeBingOptimization(html, url) {
     });
   }
   const hasLocalBusinessSchema = bingPlacesDetected;
-  const hasAddressOnPage = $("address").length > 0 || html.toLowerCase().includes("street") || /\d{5}(-\d{4})?/.test(html);
+  const hasAddressOnPage = $("address").text().trim().length > 0 || detectAddresses(html).length > 0;
   if (hasAddressOnPage && !hasLocalBusinessSchema) {
     detectedIssues.push("Local business may not be on Bing Places");
     issues.push({
@@ -12756,23 +12744,6 @@ function analyzeColorContrast(html, url) {
       });
     }
   }
-  if (cannotDetermine > elementsAnalyzed * 0.8) {
-    issues.push({
-      code: "A11Y_CONTRAST_UNDETERMINED",
-      severity: "notice",
-      category: "accessibility",
-      title: "Color contrast could not be fully analyzed",
-      description: "Most text colors are set via CSS stylesheets rather than inline styles. Full contrast checking requires browser rendering.",
-      impact: "Potential contrast issues may exist that were not detected.",
-      howToFix: "Use browser DevTools (Lighthouse accessibility audit) or dedicated tools like axe-core for comprehensive contrast checking.",
-      affectedUrls: [url],
-      details: {
-        elementsAnalyzed,
-        cannotDetermine,
-        recommendation: "Run Lighthouse accessibility audit for complete contrast analysis."
-      }
-    });
-  }
   return {
     issues,
     data: {
@@ -14094,7 +14065,8 @@ async function analyzeTrackingVerification(html, url) {
   if (!hasGa4 && !hasGtm && otherAnalytics.length === 0) {
     issues.push({
       code: "GA4_MISSING",
-      severity: "warning",
+      // A notice: which analytics a site runs is a choice, not an SEO defect.
+      severity: "notice",
       category: "on-page",
       title: "Google Analytics 4 not detected",
       description: "No GA4 tracking code found. GA4 is essential for measuring organic traffic and conversions.",
@@ -14107,7 +14079,8 @@ async function analyzeTrackingVerification(html, url) {
     const dnsChecked = verification.dnsChecked;
     issues.push({
       code: "GSC_VERIFICATION_MISSING",
-      severity: "warning",
+      // A notice: a site verified by DNS, Google Analytics or Tag Manager leaves no trace here.
+      severity: "notice",
       category: "on-page",
       title: dnsChecked ? "Google Search Console verification not found" : "Google Search Console verification not visible in the HTML",
       description: dnsChecked ? "No google-site-verification meta tag in the HTML and no google-site-verification TXT record on the domain. GSC provides essential data about your search performance." : "No google-site-verification meta tag in the HTML. The site may be verified through a DNS TXT record instead, which this run could not check. GSC provides essential data about your search performance.",
@@ -14182,6 +14155,113 @@ async function analyzeTrackingVerification(html, url) {
   return { issues, data };
 }
 
+// src/audit/meta-refresh.ts
+import * as cheerio37 from "cheerio";
+function metaRefreshTarget(html, pageUrl) {
+  if (!/http-equiv\s*=\s*["']?refresh/i.test(html)) return null;
+  const $ = cheerio37.load(html);
+  for (const el of $("meta[http-equiv]").toArray()) {
+    if (($(el).attr("http-equiv") || "").trim().toLowerCase() !== "refresh") continue;
+    const content = $(el).attr("content") || "";
+    const match = content.match(/^\s*(\d+(?:\.\d+)?)\s*[;,]\s*(?:url\s*=\s*)?['"]?\s*([^'"\s>]+)/i);
+    if (!match) continue;
+    if (Number(match[1]) > 5) continue;
+    try {
+      const target = new URL(match[2], pageUrl);
+      if (target.protocol !== "http:" && target.protocol !== "https:") continue;
+      const strip = (u) => u.origin + (u.pathname.replace(/\/+$/, "") || "/") + u.search;
+      if (strip(target) === strip(new URL(pageUrl))) continue;
+      return target.toString();
+    } catch {
+      continue;
+    }
+  }
+  return null;
+}
+
+// src/audit/dedupe.ts
+var REDUNDANT_WHEN_PRESENT = [
+  // The same tag, reported by three checks.
+  { keep: "VIEWPORT_MISSING", drop: ["HTML_NO_VIEWPORT", "RESPONSIVE_NO_VIEWPORT"] },
+  { keep: "LANG_ATTR_MISSING", drop: ["HTML_NO_LANG"] },
+  { keep: "APPLE_TOUCH_ICON_MISSING", drop: ["HTML_NO_APPLE_TOUCH_ICON"] },
+  { keep: "LLMS_TXT_MISSING", drop: ["AI_NO_LLMS_TXT"] },
+  { keep: "INDEXNOW_NOT_IMPLEMENTED", drop: ["BING_NO_INDEXNOW"] },
+  { keep: "BING_VERIFICATION_MISSING", drop: ["BING_NOT_VERIFIED"] },
+  { keep: "SITEMAP_MISSING", drop: ["BING_SITEMAP_MISSING", "SITEMAP_NOT_IN_ROBOTS"] },
+  { keep: "ROBOTS_TXT_MISSING", drop: ["SITEMAP_NOT_IN_ROBOTS"] },
+  { keep: "IMAGES_MISSING_DIMENSIONS", drop: ["IMG_NO_DIMENSIONS"] },
+  { keep: "NO_MODERN_IMAGE_FORMATS", drop: ["IMAGE_NOT_WEBP_AVIF"] },
+  { keep: "SOCIAL_PROFILES_NOT_LINKED", drop: ["NO_SOCIAL_PROFILES"] },
+  { keep: "NO_EXPERIENCE_SIGNALS", drop: ["AI_NO_EXPERIENCE_SIGNALS"] },
+  // No structured data at all makes every "your schema lacks X" finding moot.
+  {
+    keep: "SCHEMA_MISSING",
+    drop: [
+      "SCHEMA_ORG_MISSING",
+      "NO_ORGANIZATION_SCHEMA",
+      "ORGANIZATION_SCHEMA_MISSING",
+      "NO_ENTITY_SCHEMA",
+      "NO_DATE_MODIFIED_SCHEMA",
+      "NO_MAIN_ENTITY_OF_PAGE",
+      "AUTHOR_NO_SCHEMA",
+      "AI_NO_REVIEW_SCHEMA",
+      "NO_SCHEMA_SAMEAS"
+    ]
+  },
+  { keep: "NO_ORGANIZATION_SCHEMA", drop: ["ORGANIZATION_SCHEMA_MISSING"] },
+  // No title, no H1: the findings about their wording are implied.
+  { keep: "TITLE_MISSING", drop: ["LOW_CTR_POTENTIAL", "NO_KEYWORDS_IN_TITLE", "TITLE_TOO_SHORT"] },
+  { keep: "META_DESC_MISSING", drop: ["LOW_CTR_POTENTIAL"] },
+  { keep: "H1_MISSING", drop: ["NO_VISIBLE_HEADLINE", "NO_KEYWORDS_IN_H1", "H1_MISSING_KEYWORD"] },
+  // Client-side rendering, diagnosed once. CLIENT_SIDE_RENDERING carries the per-framework fix.
+  {
+    keep: "CLIENT_SIDE_RENDERING",
+    drop: [
+      "REACT_CSR_DETECTED",
+      "VUE_CSR_DETECTED",
+      "ANGULAR_NO_UNIVERSAL",
+      "SPA_WITHOUT_SSR",
+      "EMPTY_BODY_CONTENT",
+      "HIGH_JS_RENDERING_RATIO",
+      "AI_LOW_HTML_READABILITY"
+    ]
+  },
+  // An empty body has no vocabulary to be undiverse.
+  { keep: "EMPTY_BODY_CONTENT", drop: ["THIN_CONTENT", "LOW_CONTENT_ENTROPY", "HIGH_CONTENT_REPETITION", "LOW_DWELL_TIME_CONTENT", "LOW_TEXT_HTML_RATIO"] },
+  { keep: "THIN_CONTENT", drop: ["LOW_DWELL_TIME_CONTENT"] },
+  { keep: "NO_INTERNAL_LINKS", drop: ["NO_CONTENT_INTERNAL_LINKS", "HIGH_ORPHAN_RISK", "LINKS_NO_INTERNAL"] },
+  // No testimonials on the page, said once instead of six times.
+  { keep: "NO_SOCIAL_PROOF", drop: ["BOFU_NO_SOCIAL_PROOF", "AI_NO_TESTIMONIALS", "AI_NO_REVIEW_INTEGRATION", "AI_NO_REVIEW_PLATFORMS", "AI_WEAK_REVIEW_ECOSYSTEM"] },
+  { keep: "AI_NO_TESTIMONIALS", drop: ["AI_NO_REVIEW_INTEGRATION", "AI_NO_REVIEW_PLATFORMS", "AI_WEAK_REVIEW_ECOSYSTEM"] },
+  // The brand-score composite restates the brand-definition finding it is built from.
+  { keep: "AI_NO_BRAND_DEFINITION", drop: ["AI_LOW_BRAND_SCORE", "NO_ENTITY_DEFINITION"] }
+];
+function dedupeIssues(issues) {
+  const present = new Set(issues.map((i) => i.code));
+  const redundant = /* @__PURE__ */ new Set();
+  for (const rule of REDUNDANT_WHEN_PRESENT) {
+    const fires = typeof rule.keep === "string" ? present.has(rule.keep) : [...present].some((code) => rule.keep.test(code));
+    if (!fires) continue;
+    for (const code of rule.drop) redundant.add(code);
+  }
+  const kept = /* @__PURE__ */ new Map();
+  const out = [];
+  for (const issue of issues) {
+    if (redundant.has(issue.code)) continue;
+    const earlier = kept.get(issue.code);
+    if (earlier) {
+      const urls = /* @__PURE__ */ new Set([...earlier.affectedUrls || [], ...issue.affectedUrls || []]);
+      earlier.affectedUrls = [...urls];
+      continue;
+    }
+    const copy = { ...issue };
+    kept.set(issue.code, copy);
+    out.push(copy);
+  }
+  return out;
+}
+
 // src/audit/engine.ts
 function sameUrl(a, b) {
   try {
@@ -14194,15 +14274,15 @@ function sameUrl(a, b) {
   }
 }
 async function runFullAudit(options) {
-  const { url, checkBrokenLinks = false, checkHreflangUrls = false, checkCanonicalChain = false, cruxApiKey, checksLimit = 280, premiumChecks } = options;
+  const { checkBrokenLinks = false, checkHreflangUrls = false, checkCanonicalChain = false, cruxApiKey, checksLimit = 280, premiumChecks } = options;
+  let url = options.url;
   const auditRoot = isSiteRoot(url) || sameUrl(url, options.auditRootUrl ?? url);
   const allIssues = [];
   const pages = [];
   const tier = checksLimit <= 50 ? "free" : checksLimit <= 100 ? "registered" : "paid";
   const runPremiumChecks = tier === "paid" && !!premiumChecks;
   const runExtendedChecks = tier !== "free";
-  const parsedUrl = new URL(url);
-  const domain = parsedUrl.hostname;
+  let domain = new URL(url).hostname;
   console.error(`
 \u{1F50D} Running comprehensive SEO audit on ${url}...
 `);
@@ -14229,7 +14309,9 @@ async function runFullAudit(options) {
       timeout: 3e4,
       validateStatus: () => true
     }).catch((err) => {
-      console.error("Main page fetch failed:", err instanceof Error ? { message: err.message, cause: err.cause } : err);
+      const cause = err instanceof Error ? err.cause : void 0;
+      const causeText = cause instanceof Error ? ` (${cause.code || cause.message})` : "";
+      console.error(`Main page fetch failed: ${err instanceof Error ? err.message : String(err)}${causeText}`);
       return { error: err, data: "", headers: {} };
     })
   ]);
@@ -14252,7 +14334,7 @@ async function runFullAudit(options) {
     });
     return createReport(url, domain, allIssues, pages, { kind: "unreachable" });
   }
-  const statusCode = fetchResult.status;
+  let statusCode = fetchResult.status;
   if (statusCode >= 400) {
     const blocked = statusCode === 401 || statusCode === 403 || statusCode === 429;
     allIssues.push({
@@ -14265,10 +14347,42 @@ async function runFullAudit(options) {
     pages.push({ url, statusCode, wordCount: 0, loadTime: 0, issues: allIssues.map((i) => i.code) });
     return createReport(url, domain, allIssues, pages, { kind: "http-error", statusCode });
   }
-  const body = typeof fetchResult.data === "string" ? fetchResult.data : JSON.stringify(fetchResult.data ?? "");
+  const asText = (data) => typeof data === "string" ? data : JSON.stringify(data ?? "");
+  let body = asText(fetchResult.data);
+  let headers = fetchResult.headers;
+  const requestedUrl = url;
+  const refreshHops = [];
+  for (let hop = 0; hop < 3; hop++) {
+    const target = metaRefreshTarget(body, url);
+    if (!target) break;
+    try {
+      const next = await httpGet(target, { timeout: 3e4, validateStatus: () => true });
+      if (next.status >= 400) break;
+      refreshHops.push(target);
+      url = target;
+      body = asText(next.data);
+      headers = next.headers;
+      statusCode = next.status;
+    } catch {
+      break;
+    }
+  }
+  if (refreshHops.length > 0) {
+    domain = new URL(url).hostname;
+    allIssues.push({
+      code: "META_REFRESH_REDIRECT",
+      severity: "notice",
+      category: "crawlability",
+      title: "Page redirects with a meta refresh",
+      description: `${requestedUrl} answers HTTP 200 with a <meta http-equiv="refresh"> to ${url}, so that page was audited. Search engines follow an instant meta refresh, but a server redirect is the documented way to move a URL and the only one every crawler and tool honours.`,
+      impact: "Some crawlers and link-checking tools stop at the stub; link signals to the original URL may not all pass to the target.",
+      howToFix: "Return a 301 (or 308) redirect from the server or CDN for this URL instead of a meta refresh.",
+      affectedUrls: [requestedUrl],
+      details: { from: requestedUrl, to: url, hops: refreshHops }
+    });
+  }
   const html = body.replace(/<\/?noscript[^>]*>/gi, "");
-  const headers = fetchResult.headers;
-  console.error(`\u{1F4DD} Phase 2: Running synchronous HTML checks (tier: ${tier}, limit: ${checksLimit})...`);
+  console.error("\u{1F4DD} Phase 2: Running HTML checks...");
   const onPageResult = analyzeOnPage(html, url, headers, { auditRoot });
   const structuredDataResult = analyzeStructuredData2(html, url);
   const mobileResult = analyzeMobile(html, url);
@@ -14435,10 +14549,10 @@ async function runFullAudit(options) {
     if (cruxApiKey) {
       asyncChecks.push(safeAsyncIssues("Core Web Vitals", async () => {
         const cwv = await fetchCoreWebVitals(url, cruxApiKey);
-        const issues = [];
+        const issues2 = [];
         if (cwv) {
           if (cwv.lcp?.rating === "poor") {
-            issues.push({
+            issues2.push({
               code: "LCP_POOR",
               severity: "error",
               category: "performance",
@@ -14451,7 +14565,7 @@ async function runFullAudit(options) {
             });
           }
           if (cwv.cls?.rating === "poor") {
-            issues.push({
+            issues2.push({
               code: "CLS_POOR",
               severity: "error",
               category: "performance",
@@ -14464,14 +14578,15 @@ async function runFullAudit(options) {
             });
           }
         }
-        return issues;
+        return issues2;
       }));
     }
   }
   const asyncResults = await Promise.all(asyncChecks);
-  for (const issues of asyncResults) {
-    allIssues.push(...issues);
+  for (const issues2 of asyncResults) {
+    allIssues.push(...issues2);
   }
+  const issues = dedupeIssues(allIssues);
   pages.push({
     url,
     statusCode,
@@ -14481,10 +14596,10 @@ async function runFullAudit(options) {
     h1: onPageResult.data.h1s,
     wordCount: onPageResult.data.wordCount,
     loadTime: perfData.loadTime,
-    issues: allIssues.map((i) => i.code)
+    issues: issues.map((i) => i.code)
   });
   console.error("\n\u2705 Audit complete!\n");
-  return createReport(url, domain, allIssues, pages);
+  return createReport(url, domain, issues, pages);
 }
 function createReport(url, domain, issues, pages, pageFailure) {
   const healthScore = calculateHealthScore(issues);
@@ -14649,10 +14764,10 @@ function groupIssuesByCategory(issues) {
 }
 
 // src/audit/checks/duplicate-content.ts
-import * as cheerio37 from "cheerio";
+import * as cheerio38 from "cheerio";
 import { createHash } from "crypto";
 function extractContentHash(html, url) {
-  const $ = cheerio37.load(html);
+  const $ = cheerio38.load(html);
   $("script, style, nav, header, footer, aside, .nav, .header, .footer, .sidebar").remove();
   const title = $("title").text().trim();
   const bodyText = $("body").text().replace(/\s+/g, " ").trim();
@@ -15880,7 +15995,7 @@ var PRIORITY_WEIGHTS = {
 };
 
 // src/keywords/engine.ts
-import * as cheerio39 from "cheerio";
+import * as cheerio40 from "cheerio";
 
 // src/keywords/prioritizer.ts
 function prioritizeKeywords(keywords, siteProfile, existingMeta) {
@@ -16174,7 +16289,7 @@ function enrichKeywordsWithEstimates(keywords) {
 }
 
 // src/keywords/sources/free-sources.ts
-import * as cheerio38 from "cheerio";
+import * as cheerio39 from "cheerio";
 var USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36";
 async function getPeopleAlsoAsk(query) {
   try {
@@ -16183,7 +16298,7 @@ async function getPeopleAlsoAsk(query) {
       headers: { "User-Agent": USER_AGENT },
       timeout: 1e4
     });
-    const $ = cheerio38.load(response.data);
+    const $ = cheerio39.load(response.data);
     const questions = [];
     $("[data-sgrd]").each((_, el) => {
       const text = $(el).text().trim();
@@ -16209,7 +16324,7 @@ async function getRelatedSearches(query) {
       headers: { "User-Agent": USER_AGENT },
       timeout: 1e4
     });
-    const $ = cheerio38.load(response.data);
+    const $ = cheerio39.load(response.data);
     const related = [];
     $("div[data-ved] a").each((_, el) => {
       const href = $(el).attr("href");
@@ -16299,7 +16414,7 @@ async function analyzeCompetitorTitles(query) {
       headers: { "User-Agent": USER_AGENT },
       timeout: 1e4
     });
-    const $ = cheerio38.load(response.data);
+    const $ = cheerio39.load(response.data);
     const titles = [];
     const keywords = /* @__PURE__ */ new Set();
     $("h3").each((_, el) => {
@@ -16631,7 +16746,7 @@ async function fetchPageMeta(url) {
     const response = await httpGet(url, {
       timeout: 1e4
     });
-    const $ = cheerio39.load(response.data);
+    const $ = cheerio40.load(response.data);
     return {
       url,
       title: $("title").text().trim() || void 0,
@@ -16762,20 +16877,17 @@ async function extractSeedKeywords(url) {
     const response = await httpGet(url, {
       timeout: 1e4
     });
-    const $ = cheerio39.load(response.data);
+    const $ = cheerio40.load(response.data);
     const seeds = /* @__PURE__ */ new Set();
-    const title = $("title").text().toLowerCase();
-    const titleWords = title.split(/[\s\-|:]+/).filter((w) => w.length > 3);
-    titleWords.forEach((w) => seeds.add(w));
-    const h1 = $("h1").first().text().toLowerCase();
-    const h1Words = h1.split(/[\s\-|:]+/).filter((w) => w.length > 3);
-    h1Words.forEach((w) => seeds.add(w));
+    const words = (text) => text.toLowerCase().split(/[\s\-|:]+/).map((w) => w.replace(/^[^\p{L}\p{N}]+|[^\p{L}\p{N}]+$/gu, "")).filter((w) => w.length > 3);
+    words($("title").text()).forEach((w) => seeds.add(w));
+    words($("h1").first().text()).forEach((w) => seeds.add(w));
     const metaKeywords = $('meta[name="keywords"]').attr("content");
     if (metaKeywords) {
       metaKeywords.split(",").forEach((k) => seeds.add(k.trim().toLowerCase()));
     }
     const desc = $('meta[name="description"]').attr("content")?.toLowerCase() || "";
-    const descWords = desc.split(/\s+/).filter((w) => w.length > 5);
+    const descWords = words(desc).filter((w) => w.length > 5);
     descWords.slice(0, 5).forEach((w) => seeds.add(w));
     return Array.from(seeds).slice(0, 10);
   } catch {
@@ -16828,7 +16940,7 @@ function parseJsonReply(text) {
 }
 
 // src/keywords/site-crawler.ts
-import * as cheerio40 from "cheerio";
+import * as cheerio41 from "cheerio";
 var EXCLUDED_PATHS = [
   "/cdn-cgi/",
   "/wp-admin/",
@@ -16940,7 +17052,7 @@ async function crawlPage(url, timeout) {
     validateStatus: (status) => status === 200
   });
   const html = response.data;
-  const $ = cheerio40.load(html);
+  const $ = cheerio41.load(html);
   $('script, style, noscript, iframe, nav, footer, header, aside, [role="navigation"]').remove();
   const title = $("title").text().trim();
   const description = $('meta[name="description"]').attr("content")?.trim() || "";
@@ -19575,7 +19687,7 @@ function getDateRange(days = 28) {
 }
 
 // src/keywords/sources/competitor-analysis.ts
-import * as cheerio41 from "cheerio";
+import * as cheerio42 from "cheerio";
 var USER_AGENT2 = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36";
 async function discoverCompetitorKeywords(yourDomain, seedKeywords, competitors) {
   const yourKeywords = /* @__PURE__ */ new Set();
@@ -19672,7 +19784,7 @@ async function analyzeSERP(query) {
       headers: { "User-Agent": USER_AGENT2 },
       timeout: 1e4
     });
-    const $ = cheerio41.load(response.data);
+    const $ = cheerio42.load(response.data);
     const results = [];
     const relatedSearches = [];
     const peopleAlsoAsk = [];
@@ -20125,7 +20237,7 @@ function formatTopicReport(result) {
 
 // src/content/readability.ts
 function analyzeReadability(text) {
-  const cleanText = text.replace(/<[^>]*>/g, " ").replace(/\s+/g, " ").trim();
+  const cleanText = text.replace(/<[^>]*>/g, " ").replace(/[ \t\r\f\v]+/g, " ").replace(/ ?\n ?/g, "\n").replace(/\n{3,}/g, "\n\n").trim();
   const words = getWords(cleanText);
   const sentences = getSentences(cleanText);
   const paragraphs = getParagraphs(cleanText);
@@ -20194,7 +20306,7 @@ function getSentences(text) {
   return text.split(/[.!?]+/).filter((s) => s.trim().length > 0);
 }
 function getParagraphs(text) {
-  return text.split(/\n\n+/).filter((p) => p.trim().length > 0);
+  return text.split(/\n\n+/).filter((p) => p.trim().split(/\s+/).length >= 3);
 }
 function countSyllables(word) {
   word = word.toLowerCase().replace(/[^a-z]/g, "");
@@ -32462,6 +32574,7 @@ export {
   LOCATION_CODES,
   OG_IMAGE_SPECS,
   PRIORITY_WEIGHTS,
+  REDUNDANT_WHEN_PRESENT,
   RankTracker,
   SEO_SCOPES,
   SITE_PROFILE_QUESTIONS,
@@ -32545,6 +32658,7 @@ export {
   createPullRequest,
   createSEOCommit,
   createSEOCommits,
+  dedupeIssues,
   describeConcerns,
   detectDuplicates,
   detectFramework4 as detectFramework,
@@ -32681,6 +32795,7 @@ export {
   loadAgent,
   loadAgentByName,
   mergePages,
+  metaRefreshTarget,
   openAIKeywordLLM,
   optimizeForAI,
   optimizeReadme,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@rankcli/agent-runtime",
-  "version": "0.0.38",
+  "version": "0.0.39",
   "description": "RankCLI agent runtime - executes SEO audits and fixes with AI",
   "homepage": "https://rankcli.dev",
   "main": "dist/index.js",