@houtini/seo-audit-console 0.2.0

package/dist/core/extract.js

@@ -0,0 +1,262 @@
+import * as cheerio from 'cheerio';
+import { urlKey } from './url-key.js';
+/**
+ * Internal = the seed host or its www/apex twin ONLY — not arbitrary sibling subdomains.
+ * accounts./shop./blog. etc. are separate properties; crawling (or following redirects into)
+ * them pulls in junk like Shopify's accounts.<domain> OAuth login pages. Shared by the
+ * link-classifier here and the crawler's redirect-target guard.
+ */
+export const isInternalHost = (targetHost, baseHost) => {
+    const apex = (h) => (h.startsWith('www.') ? h.slice(4) : h);
+    const a = apex(baseHost.toLowerCase());
+    const t = targetHost.toLowerCase();
+    return t === a || t === `www.${a}`;
+};
+function placementOf($, el) {
+    const $el = $(el);
+    if ($el.closest('nav, header').length)
+        return 'navigation';
+    if ($el.closest('footer').length)
+        return 'footer';
+    return 'body';
+}
+/**
+ * Extract the SEO-relevant surface from raw HTML. Pure function of bytes (D).
+ * `baseHost` is the seed apex host for internal/external classification;
+ * `keyOpts` is the url-key normalisation config for this property.
+ */
+export function extractPage(html, pageUrl, baseHost, keyOpts, xRobotsTag) {
+    const $ = cheerio.load(html);
+    const title = $('head > title').first().text().trim() || null;
+    const metaDescription = $('meta[name="description"]').attr('content')?.trim() ?? null;
+    const h1Els = $('h1');
+    const h1 = h1Els.first().text().trim() || null;
+    const lang = $('html').attr('lang')?.trim() ?? null;
+    const charset = $('meta[charset]').attr('charset')?.trim()
+        ?? ($('meta[http-equiv="Content-Type"]').attr('content')?.match(/charset=([\w-]+)/i)?.[1] ?? null);
+    const canonicalEls = $('link[rel="canonical"]');
+    const canonicalRaw = canonicalEls.first().attr('href')?.trim() ?? null;
+    // A malformed canonical href (e.g. `href="http://"`) must not make the whole page
+    // extraction throw — that would record a fine 200 page as a fetch failure.
+    const canonicalUrl = canonicalRaw ? (() => { try {
+        return new URL(canonicalRaw, pageUrl).toString();
+    }
+    catch {
+        return null;
+    } })() : null;
+    const canonicalRelative = canonicalRaw != null && !/^https?:\/\//i.test(canonicalRaw);
+    const robots = $('meta[name="robots"]').attr('content')?.trim() ?? null;
+    const viewport = $('meta[name="viewport"]').attr('content')?.trim() ?? null;
+    const noindex = /noindex/i.test(robots ?? '') || /noindex/i.test(xRobotsTag ?? '');
+    // word count excluding script/style/noscript/template (fixes inline-script inflation)
+    const $body = cheerio.load(html);
+    $body('script, style, noscript, template').remove();
+    const bodyText = $body('body').text().replace(/\s+/g, ' ').trim();
+    const wordCount = bodyText ? bodyText.split(' ').length : 0;
+    // Body chunks: visible content segmented by heading, boilerplate stripped — the data layer for the
+    // content-cluster checks (phrase gaps, RAG alignment, extractability). Depth-first reading-order
+    // walk: a heading starts a new chunk; other text accrues to the current one. No DOM mutation.
+    const bodyChunks = [];
+    {
+        const $c = cheerio.load(html);
+        $c('script, style, noscript, template, nav, header, footer, aside').remove();
+        const root = ($c('main').first()[0] ?? $c('article').first()[0] ?? $c('body')[0]);
+        const ws = (t) => t.replace(/\s+/g, ' ').trim();
+        let cur = { heading: '', level: 0, parts: [] };
+        const flush = () => {
+            const text = ws(cur.parts.join(' '));
+            if (cur.heading || text)
+                bodyChunks.push({ heading: cur.heading, level: cur.level, text: text.slice(0, 1500) });
+        };
+        const walk = (node) => {
+            for (const child of node.children ?? []) {
+                if (child.type === 'tag') {
+                    if (/^h[1-6]$/.test(child.name)) {
+                        flush();
+                        cur = { heading: ws($c(child).text()), level: Number(child.name[1]), parts: [] };
+                    }
+                    else {
+                        walk(child);
+                    }
+                }
+                else if (child.type === 'text' && child.data && child.data.trim()) {
+                    cur.parts.push(child.data);
+                }
+            }
+        };
+        if (root) {
+            walk(root);
+            flush();
+        }
+        const kept = bodyChunks.filter(c => c.heading || c.text.length > 20).slice(0, 40);
+        bodyChunks.length = 0;
+        bodyChunks.push(...kept);
+    }
+    const jsonLdBlocks = [];
+    $('script[type="application/ld+json"]').each((_, el) => {
+        const t = $(el).contents().text().trim();
+        if (t)
+            jsonLdBlocks.push(t);
+    });
+    const og = {};
+    $('meta[property^="og:"]').each((_, el) => {
+        const p = $(el).attr('property');
+        const c = $(el).attr('content');
+        if (p && c)
+            og[p] = c;
+    });
+    const hreflang = [];
+    $('link[rel="alternate"][hreflang]').each((_, el) => {
+        const l = $(el).attr('hreflang');
+        const href = $(el).attr('href');
+        if (l && href)
+            hreflang.push({ lang: l, href });
+    });
+    // Twitter Card tags (alongside Open Graph) — share-appearance completeness.
+    const tw = {};
+    $('meta[name^="twitter:"]').each((_, el) => {
+        const n = $(el).attr('name');
+        const c = $(el).attr('content');
+        if (n && c)
+            tw[n] = c;
+    });
+    // Heading hierarchy: levels in document order → count skipped jumps (h1→h3 etc.).
+    const headingLevels = [];
+    $('h1, h2, h3, h4, h5, h6').each((_, el) => {
+        const lvl = Number(el.name?.[1]);
+        if (lvl >= 1 && lvl <= 6)
+            headingLevels.push(lvl);
+    });
+    let headingSkips = 0;
+    for (let i = 1; i < headingLevels.length; i++)
+        if (headingLevels[i] - headingLevels[i - 1] > 1)
+            headingSkips++;
+    const h2Count = headingLevels.filter(l => l === 2).length;
+    // rel link relations Google still reads / we don't otherwise capture.
+    const relNext = $('link[rel="next"]').length > 0;
+    const relPrev = $('link[rel="prev"]').length > 0;
+    const relAmphtml = $('link[rel="amphtml"]').attr('href')?.trim() ?? null;
+    // Mixed content: insecure (http) subresources on an https page.
+    const isHttps = /^https:/i.test(pageUrl);
+    let mixedContentCount = 0;
+    if (isHttps) {
+        $('img[src], script[src], link[rel="stylesheet"][href], iframe[src], source[src], video[src], audio[src]').each((_, el) => {
+            const u = $(el).attr('src') || $(el).attr('href') || '';
+            if (/^http:\/\//i.test(u))
+                mixedContentCount++;
+        });
+    }
+    // Structured-data formats beyond JSON-LD (RDFa via typeof/vocab, not property — OG meta uses property).
+    const hasMicrodata = $('[itemscope]').length > 0;
+    const hasRdfa = $('[typeof], [vocab]').length > 0;
+    // Favicon: any icon link relation (icon / shortcut icon / apple-touch-icon).
+    const hasFavicon = $('link[rel~="icon"], link[rel="shortcut icon"], link[rel="apple-touch-icon"]').length > 0;
+    // Client-side analytics: a known snippet/host anywhere in the raw HTML. Deliberately a tight
+    // allow-list of real analytics products — absence means "no CLIENT-SIDE analytics detected",
+    // not "no analytics" (server-side measurement is invisible to a crawl).
+    const hasAnalytics = /googletagmanager\.com|google-analytics\.com|gtag\(|plausible\.io|matomo|_paq\b|usefathom\.com|clarity\.ms|cloudflareinsights\.com|mixpanel|heap-\d|segment\.com\/analytics/i.test(html);
+    // Images: count those lacking a non-empty alt (decorative alt="" is intentional, not flagged),
+    // and those lacking both width & height attributes (a deterministic CLS signal).
+    const imgEls = $('img');
+    let imagesWithoutAlt = 0;
+    let imagesMissingDimensions = 0;
+    // Distinct absolute image URLs, capped per page — the crawler samples their content-length
+    // after the crawl for the image-weight report (CDN-hosted images count; SEO cares about
+    // what the page loads, not where it's hosted).
+    const imageSrcSet = new Set();
+    imgEls.each((_, el) => {
+        const $e = $(el);
+        if ($e.attr('alt') === undefined)
+            imagesWithoutAlt++;
+        if ($e.attr('width') === undefined && $e.attr('height') === undefined)
+            imagesMissingDimensions++;
+        const src = ($e.attr('src') ?? '').trim();
+        if (src && !src.startsWith('data:') && imageSrcSet.size < 15) {
+            try {
+                const u = new URL(src, pageUrl);
+                if (u.protocol === 'http:' || u.protocol === 'https:')
+                    imageSrcSet.add(u.toString());
+            }
+            catch { /* skip bad src */ }
+        }
+    });
+    const links = [];
+    let internalLinks = 0;
+    let externalLinks = 0;
+    $('a[href]').each((_, el) => {
+        const href = $(el).attr('href');
+        if (!href)
+            return;
+        let target;
+        try {
+            target = new URL(href, pageUrl);
+        }
+        catch {
+            return;
+        }
+        if (target.protocol !== 'http:' && target.protocol !== 'https:')
+            return;
+        // Cloudflare-reserved infrastructure (/cdn-cgi/l/email-protection, rocket-loader, …) —
+        // never real content and 404s to bots by design; skip so it's neither crawled nor counted
+        // as an internal link (otherwise it shows up as a high-equity "broken link" false positive).
+        if (/^\/cdn-cgi\//i.test(target.pathname))
+            return;
+        // Internal = the seed host or its www/apex twin ONLY — not arbitrary sibling subdomains.
+        // (accounts./shop./blog. etc. are separate properties; crawling them pulls in junk like
+        // Shopify's accounts.<domain> OAuth login pages — infinite param URLs, no SEO value.)
+        const isInternal = isInternalHost(target.hostname, baseHost);
+        if (isInternal)
+            internalLinks++;
+        else
+            externalLinks++;
+        links.push({
+            targetUrl: target.toString(),
+            targetKey: urlKey(target.toString(), keyOpts),
+            anchor: $(el).text().replace(/\s+/g, ' ').trim().slice(0, 300),
+            placement: placementOf($, el),
+            rel: $(el).attr('rel')?.trim() ?? '',
+            isInternal,
+        });
+    });
+    return {
+        title,
+        titleLength: title?.length ?? 0,
+        metaDescription,
+        metaDescriptionLength: metaDescription?.length ?? 0,
+        h1,
+        h1Count: h1Els.length,
+        wordCount,
+        lang,
+        charset,
+        canonicalUrl,
+        canonicalKey: canonicalUrl ? urlKey(canonicalUrl, keyOpts) : null,
+        robots,
+        viewport,
+        noindex,
+        jsonLd: jsonLdBlocks.length ? JSON.stringify(jsonLdBlocks) : null,
+        ogTags: Object.keys(og).length ? JSON.stringify(og) : null,
+        hreflang: hreflang.length ? JSON.stringify(hreflang) : null,
+        internalLinks,
+        externalLinks,
+        imageCount: imgEls.length,
+        imagesWithoutAlt,
+        imagesMissingDimensions,
+        canonicalCount: canonicalEls.length,
+        canonicalRelative,
+        h2Count,
+        headingSkips,
+        relNext,
+        relPrev,
+        relAmphtml,
+        mixedContentCount,
+        twitterTags: Object.keys(tw).length ? JSON.stringify(tw) : null,
+        hasMicrodata,
+        hasRdfa,
+        hasFavicon,
+        hasAnalytics,
+        bodyChunks,
+        links,
+        imageSrcs: [...imageSrcSet],
+    };
+}
+//# sourceMappingURL=extract.js.map

package/dist/core/extract.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"extract.js","sourceRoot":"","sources":["../../src/core/extract.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,OAAO,MAAM,SAAS,CAAC;AACnC,OAAO,EAAE,MAAM,EAAsB,MAAM,cAAc,CAAC;AAE1D;;;;;GAKG;AACH,MAAM,CAAC,MAAM,cAAc,GAAG,CAAC,UAAkB,EAAE,QAAgB,EAAW,EAAE;IAC9E,MAAM,IAAI,GAAG,CAAC,CAAS,EAAU,EAAE,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC5E,MAAM,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC,CAAC;IACvC,MAAM,CAAC,GAAG,UAAU,CAAC,WAAW,EAAE,CAAC;IACnC,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,OAAO,CAAC,EAAE,CAAC;AACrC,CAAC,CAAC;AAoDF,SAAS,WAAW,CAAC,CAAqB,EAAE,EAAO;IACjD,MAAM,GAAG,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC;IAClB,IAAI,GAAG,CAAC,OAAO,CAAC,aAAa,CAAC,CAAC,MAAM;QAAE,OAAO,YAAY,CAAC;IAC3D,IAAI,GAAG,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,MAAM;QAAE,OAAO,QAAQ,CAAC;IAClD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,WAAW,CACzB,IAAY,EACZ,OAAe,EACf,QAAgB,EAChB,OAAsB,EACtB,UAA0B;IAE1B,MAAM,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAE7B,MAAM,KAAK,GAAG,CAAC,CAAC,cAAc,CAAC,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,CAAC,IAAI,EAAE,IAAI,IAAI,CAAC;IAC9D,MAAM,eAAe,GAAG,CAAC,CAAC,0BAA0B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE,IAAI,IAAI,CAAC;IACtF,MAAM,KAAK,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC;IACtB,MAAM,EAAE,GAAG,KAAK,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,CAAC,IAAI,EAAE,IAAI,IAAI,CAAC;IAC/C,MAAM,IAAI,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,IAAI,EAAE,IAAI,IAAI,CAAC;IACpD,MAAM,OAAO,GAAG,CAAC,CAAC,eAAe,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE;WACrD,CAAC,CAAC,CAAC,iCAAiC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,KAAK,CAAC,mBAAmB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;IACrG,MAAM,YAAY,GAAG,CAAC,CAAC,uBAAuB,CAAC,CAAC;IAChD,MAAM,YAAY,GAAG,YAAY,CAAC,KAAK,EAAE,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,IAAI,EAAE,IAAI,IAAI,CAAC;IACvE,kFAAkF;IAClF,2EAA2E;IAC3E,MAAM,YAAY,GAAG,YAAY,CAAC,CAAC,CAAC,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC;QAAC,OAAO,IAAI,GAAG,CAAC,YAAY,EAAE,OAAO,CAAC,CAAC,QAAQ,EAAE,CAAC;IAAC,CAAC;IAAC,MAAM,CAAC;QAAC,OAAO,IAAI,CAAC;IAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;IAC1I,MAAM,iBAAiB,GAAG,YAAY,IAAI,IAAI,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC;IACtF,MAAM,MAAM,GAAG,CAAC,CAAC,qBAAqB,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE,IAAI,IAAI,CAAC;IACxE,MAAM,QAAQ,GAAG,CAAC,CAAC,uBAAuB,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,IAAI,EAAE,IAAI,IAAI,CAAC;IAC5E,MAAM,OAAO,GAAG,UAAU,CAAC,IAAI,CAAC,MAAM,IAAI,EAAE,CAAC,IAAI,UAAU,CAAC,IAAI,CAAC,UAAU,IAAI,EAAE,CAAC,CAAC;IAEnF,sFAAsF;IACtF,MAAM,KAAK,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACjC,KAAK,CAAC,mCAAmC,CAAC,CAAC,MAAM,EAAE,CAAC;IACpD,MAAM,QAAQ,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;IAClE,MAAM,SAAS,GAAG,QAAQ,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC;IAE5D,mGAAmG;IACnG,iGAAiG;IACjG,8FAA8F;IAC9F,MAAM,UAAU,GAAuD,EAAE,CAAC;IAC1E,CAAC;QACC,MAAM,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAC9B,EAAE,CAAC,+DAA+D,CAAC,CAAC,MAAM,EAAE,CAAC;QAC7E,MAAM,IAAI,GAAG,CAAC,EAAE,CAAC,MAAM,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,SAAS,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAQ,CAAC;QACzF,MAAM,EAAE,GAAG,CAAC,CAAS,EAAU,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;QAChE,IAAI,GAAG,GAAwD,EAAE,OAAO,EAAE,EAAE,EAAE,KAAK,EAAE,CAAC,EAAE,KAAK,EAAE,EAAE,EAAE,CAAC;QACpG,MAAM,KAAK,GAAG,GAAS,EAAE;YACvB,MAAM,IAAI,GAAG,EAAE,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;YACrC,IAAI,GAAG,CAAC,OAAO,IAAI,IAAI;gBAAE,UAAU,CAAC,IAAI,CAAC,EAAE,OAAO,EAAE,GAAG,CAAC,OAAO,EAAE,KAAK,EAAE,GAAG,CAAC,KAAK,EAAE,IAAI,EAAE,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,EAAE,CAAC,CAAC;QAClH,CAAC,CAAC;QACF,MAAM,IAAI,GAAG,CAAC,IAAS,EAAQ,EAAE;YAC/B,KAAK,MAAM,KAAK,IAAI,IAAI,CAAC,QAAQ,IAAI,EAAE,EAAE,CAAC;gBACxC,IAAI,KAAK,CAAC,IAAI,KAAK,KAAK,EAAE,CAAC;oBACzB,IAAI,UAAU,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC;wBAChC,KAAK,EAAE,CAAC;wBACR,GAAG,GAAG,EAAE,OAAO,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,CAAC,CAAC,IAAI,EAAE,CAAC,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,EAAE,EAAE,CAAC;oBACnF,CAAC;yBAAM,CAAC;wBACN,IAAI,CAAC,KAAK,CAAC,CAAC;oBACd,CAAC;gBACH,CAAC;qBAAM,IAAI,KAAK,CAAC,IAAI,KAAK,MAAM,IAAI,KAAK,CAAC,IAAI,IAAI,KAAK,CAAC,IAAI,CAAC,IAAI,EAAE,EAAE,CAAC;oBACpE,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;gBAC7B,CAAC;YACH,CAAC;QACH,CAAC,CAAC;QACF,IAAI,IAAI,EAAE,CAAC;YAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAAC,KAAK,EAAE,CAAC;QAAC,CAAC;QAClC,MAAM,IAAI,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,IAAI,CAAC,MAAM,GAAG,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QAClF,UAAU,CAAC,MAAM,GAAG,CAAC,CAAC;QACtB,UAAU,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,CAAC;IAC3B,CAAC;IAED,MAAM,YAAY,GAAa,EAAE,CAAC;IAClC,CAAC,CAAC,oCAAoC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;QACrD,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,QAAQ,EAAE,CAAC,IAAI,EAAE,CAAC,IAAI,EAAE,CAAC;QACzC,IAAI,CAAC;YAAE,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IAC9B,CAAC,CAAC,CAAC;IAEH,MAAM,EAAE,GAA2B,EAAE,CAAC;IACtC,CAAC,CAAC,uBAAuB,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;QACxC,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;QACjC,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;QAChC,IAAI,CAAC,IAAI,CAAC;YAAE,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;IACxB,CAAC,CAAC,CAAC;IAEH,MAAM,QAAQ,GAAqC,EAAE,CAAC;IACtD,CAAC,CAAC,iCAAiC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;QAClD,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;QACjC,MAAM,IAAI,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAChC,IAAI,CAAC,IAAI,IAAI;YAAE,QAAQ,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC;IAClD,CAAC,CAAC,CAAC;IAEH,4EAA4E;IAC5E,MAAM,EAAE,GAA2B,EAAE,CAAC;IACtC,CAAC,CAAC,wBAAwB,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;QACzC,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAC7B,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;QAChC,IAAI,CAAC,IAAI,CAAC;YAAE,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;IACxB,CAAC,CAAC,CAAC;IAEH,kFAAkF;IAClF,MAAM,aAAa,GAAa,EAAE,CAAC;IACnC,CAAC,CAAC,wBAAwB,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;QACzC,MAAM,GAAG,GAAG,MAAM,CAAE,EAAwB,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QACxD,IAAI,GAAG,IAAI,CAAC,IAAI,GAAG,IAAI,CAAC;YAAE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACpD,CAAC,CAAC,CAAC;IACH,IAAI,YAAY,GAAG,CAAC,CAAC;IACrB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,aAAa,CAAC,MAAM,EAAE,CAAC,EAAE;QAAE,IAAI,aAAa,CAAC,CAAC,CAAC,GAAG,aAAa,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC;YAAE,YAAY,EAAE,CAAC;IAC/G,MAAM,OAAO,GAAG,aAAa,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC;IAE1D,sEAAsE;IACtE,MAAM,OAAO,GAAG,CAAC,CAAC,kBAAkB,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;IACjD,MAAM,OAAO,GAAG,CAAC,CAAC,kBAAkB,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;IACjD,MAAM,UAAU,GAAG,CAAC,CAAC,qBAAqB,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,IAAI,EAAE,IAAI,IAAI,CAAC;IAEzE,gEAAgE;IAChE,MAAM,OAAO,GAAG,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IACzC,IAAI,iBAAiB,GAAG,CAAC,CAAC;IAC1B,IAAI,OAAO,EAAE,CAAC;QACZ,CAAC,CAAC,uGAAuG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;YACxH,MAAM,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC;YACxD,IAAI,aAAa,CAAC,IAAI,CAAC,CAAC,CAAC;gBAAE,iBAAiB,EAAE,CAAC;QACjD,CAAC,CAAC,CAAC;IACL,CAAC;IAED,wGAAwG;IACxG,MAAM,YAAY,GAAG,CAAC,CAAC,aAAa,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;IACjD,MAAM,OAAO,GAAG,CAAC,CAAC,mBAAmB,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;IAElD,6EAA6E;IAC7E,MAAM,UAAU,GAAG,CAAC,CAAC,4EAA4E,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC;IAC9G,6FAA6F;IAC7F,6FAA6F;IAC7F,wEAAwE;IACxE,MAAM,YAAY,GAAG,6KAA6K,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IAE9M,+FAA+F;IAC/F,iFAAiF;IACjF,MAAM,MAAM,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;IACxB,IAAI,gBAAgB,GAAG,CAAC,CAAC;IACzB,IAAI,uBAAuB,GAAG,CAAC,CAAC;IAChC,2FAA2F;IAC3F,wFAAwF;IACxF,+CAA+C;IAC/C,MAAM,WAAW,GAAG,IAAI,GAAG,EAAU,CAAC;IACtC,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;QACpB,MAAM,EAAE,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC;QACjB,IAAI,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,SAAS;YAAE,gBAAgB,EAAE,CAAC;QACrD,IAAI,EAAE,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,SAAS,IAAI,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,KAAK,SAAS;YAAE,uBAAuB,EAAE,CAAC;QACjG,MAAM,GAAG,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;QAC1C,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,WAAW,CAAC,IAAI,GAAG,EAAE,EAAE,CAAC;YAC7D,IAAI,CAAC;gBACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,EAAE,OAAO,CAAC,CAAC;gBAChC,IAAI,CAAC,CAAC,QAAQ,KAAK,OAAO,IAAI,CAAC,CAAC,QAAQ,KAAK,QAAQ;oBAAE,WAAW,CAAC,GAAG,CAAC,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC;YACvF,CAAC;YAAC,MAAM,CAAC,CAAC,kBAAkB,CAAC,CAAC;QAChC,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,MAAM,KAAK,GAAoB,EAAE,CAAC;IAClC,IAAI,aAAa,GAAG,CAAC,CAAC;IACtB,IAAI,aAAa,GAAG,CAAC,CAAC;IACtB,CAAC,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,EAAE;QAC1B,MAAM,IAAI,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAChC,IAAI,CAAC,IAAI;YAAE,OAAO;QAClB,IAAI,MAAW,CAAC;QAChB,IAAI,CAAC;YACH,MAAM,GAAG,IAAI,GAAG,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;QAClC,CAAC;QAAC,MAAM,CAAC;YACP,OAAO;QACT,CAAC;QACD,IAAI,MAAM,CAAC,QAAQ,KAAK,OAAO,IAAI,MAAM,CAAC,QAAQ,KAAK,QAAQ;YAAE,OAAO;QACxE,uFAAuF;QACvF,0FAA0F;QAC1F,6FAA6F;QAC7F,IAAI,eAAe,CAAC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC;YAAE,OAAO;QAClD,yFAAyF;QACzF,wFAAwF;QACxF,sFAAsF;QACtF,MAAM,UAAU,GAAG,cAAc,CAAC,MAAM,CAAC,QAAQ,EAAE,QAAQ,CAAC,CAAC;QAC7D,IAAI,UAAU;YAAE,aAAa,EAAE,CAAC;;YAC3B,aAAa,EAAE,CAAC;QACrB,KAAK,CAAC,IAAI,CAAC;YACT,SAAS,EAAE,MAAM,CAAC,QAAQ,EAAE;YAC5B,SAAS,EAAE,MAAM,CAAC,MAAM,CAAC,QAAQ,EAAE,EAAE,OAAO,CAAC;YAC7C,MAAM,EAAE,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC;YAC9D,SAAS,EAAE,WAAW,CAAC,CAAC,EAAE,EAAE,CAAC;YAC7B,GAAG,EAAE,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE;YACpC,UAAU;SACX,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,OAAO;QACL,KAAK;QACL,WAAW,EAAE,KAAK,EAAE,MAAM,IAAI,CAAC;QAC/B,eAAe;QACf,qBAAqB,EAAE,eAAe,EAAE,MAAM,IAAI,CAAC;QACnD,EAAE;QACF,OAAO,EAAE,KAAK,CAAC,MAAM;QACrB,SAAS;QACT,IAAI;QACJ,OAAO;QACP,YAAY;QACZ,YAAY,EAAE,YAAY,CAAC,CAAC,CAAC,MAAM,CAAC,YAAY,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI;QACjE,MAAM;QACN,QAAQ;QACR,OAAO;QACP,MAAM,EAAE,YAAY,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC,IAAI;QACjE,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI;QAC1D,QAAQ,EAAE,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,IAAI;QAC3D,aAAa;QACb,aAAa;QACb,UAAU,EAAE,MAAM,CAAC,MAAM;QACzB,gBAAgB;QAChB,uBAAuB;QACvB,cAAc,EAAE,YAAY,CAAC,MAAM;QACnC,iBAAiB;QACjB,OAAO;QACP,YAAY;QACZ,OAAO;QACP,OAAO;QACP,UAAU;QACV,iBAAiB;QACjB,WAAW,EAAE,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI;QAC/D,YAAY;QACZ,OAAO;QACP,UAAU;QACV,YAAY;QACZ,UAAU;QACV,KAAK;QACL,SAAS,EAAE,CAAC,GAAG,WAAW,CAAC;KAC5B,CAAC;AACJ,CAAC"}

package/dist/core/gscFreshness.d.ts

@@ -0,0 +1,18 @@
+import type Database from 'better-sqlite3';
+export interface GscFreshness {
+    /** Latest date present in search_analytics (may be a partial-data day). */
+    rawMax: string | null;
+    /** Latest date whose data looks finalised — use this for windows and charts. */
+    effectiveMax: string | null;
+    /** How many trailing days were trimmed as still-being-finalised. */
+    trimmedDays: number;
+}
+/**
+ * Google Search Console finalises data ~2–3 days late, so the most recent day(s) in the DB are
+ * partial — a sharp downward cliff that reads as "traffic is tanking" and panics SEOs. Detect and
+ * trim those trailing partial days *data-drivenly*: drop any trailing day whose impressions fall
+ * below 50% of the recent (stable-window) median, capped at 3 days. This catches the crater on
+ * busy sites without over-trimming a genuinely low-traffic site whose every day is small.
+ */
+export declare function gscFreshness(db: Database.Database): GscFreshness;
+//# sourceMappingURL=gscFreshness.d.ts.map

package/dist/core/gscFreshness.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"gscFreshness.d.ts","sourceRoot":"","sources":["../../src/core/gscFreshness.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,QAAQ,MAAM,gBAAgB,CAAC;AAE3C,MAAM,WAAW,YAAY;IAC3B,2EAA2E;IAC3E,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;IACtB,gFAAgF;IAChF,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,oEAAoE;IACpE,WAAW,EAAE,MAAM,CAAC;CACrB;AAED;;;;;;GAMG;AACH,wBAAgB,YAAY,CAAC,EAAE,EAAE,QAAQ,CAAC,QAAQ,GAAG,YAAY,CAoBhE"}

package/dist/core/gscFreshness.js

@@ -0,0 +1,32 @@
+/**
+ * Google Search Console finalises data ~2–3 days late, so the most recent day(s) in the DB are
+ * partial — a sharp downward cliff that reads as "traffic is tanking" and panics SEOs. Detect and
+ * trim those trailing partial days *data-drivenly*: drop any trailing day whose impressions fall
+ * below 50% of the recent (stable-window) median, capped at 3 days. This catches the crater on
+ * busy sites without over-trimming a genuinely low-traffic site whose every day is small.
+ */
+export function gscFreshness(db) {
+    const rows = db
+        .prepare(`SELECT date, SUM(impressions) i FROM search_analytics WHERE date IS NOT NULL GROUP BY date ORDER BY date`)
+        .all();
+    if (rows.length === 0)
+        return { rawMax: null, effectiveMax: null, trimmedDays: 0 };
+    const rawMax = rows[rows.length - 1].date;
+    if (rows.length < 8)
+        return { rawMax, effectiveMax: rawMax, trimmedDays: 0 };
+    // Reference level from a recent stable window (last ~21 days, excluding the last 3 that may be partial).
+    const stable = rows.slice(Math.max(0, rows.length - 24), rows.length - 3).map(r => r.i).sort((a, b) => a - b);
+    const median = stable.length ? stable[Math.floor(stable.length / 2)] : 0;
+    if (median <= 0)
+        return { rawMax, effectiveMax: rawMax, trimmedDays: 0 };
+    let trimmed = 0;
+    for (let k = rows.length - 1; k >= 0 && trimmed < 3; k--) {
+        if (rows[k].i < median * 0.5)
+            trimmed++;
+        else
+            break;
+    }
+    const effectiveMax = rows[rows.length - 1 - trimmed].date;
+    return { rawMax, effectiveMax, trimmedDays: trimmed };
+}
+//# sourceMappingURL=gscFreshness.js.map

package/dist/core/gscFreshness.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"gscFreshness.js","sourceRoot":"","sources":["../../src/core/gscFreshness.ts"],"names":[],"mappings":"AAWA;;;;;;GAMG;AACH,MAAM,UAAU,YAAY,CAAC,EAAqB;IAChD,MAAM,IAAI,GAAG,EAAE;SACZ,OAAO,CAAC,0GAA0G,CAAC;SACnH,GAAG,EAAmC,CAAC;IAC1C,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,MAAM,EAAE,IAAI,EAAE,YAAY,EAAE,IAAI,EAAE,WAAW,EAAE,CAAC,EAAE,CAAC;IACnF,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC;IAC1C,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,EAAE,MAAM,EAAE,YAAY,EAAE,MAAM,EAAE,WAAW,EAAE,CAAC,EAAE,CAAC;IAE7E,yGAAyG;IACzG,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,MAAM,GAAG,EAAE,CAAC,EAAE,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IAC9G,MAAM,MAAM,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACzE,IAAI,MAAM,IAAI,CAAC;QAAE,OAAO,EAAE,MAAM,EAAE,YAAY,EAAE,MAAM,EAAE,WAAW,EAAE,CAAC,EAAE,CAAC;IAEzE,IAAI,OAAO,GAAG,CAAC,CAAC;IAChB,KAAK,IAAI,CAAC,GAAG,IAAI,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,IAAI,OAAO,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QACzD,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,MAAM,GAAG,GAAG;YAAE,OAAO,EAAE,CAAC;;YACnC,MAAM;IACb,CAAC;IACD,MAAM,YAAY,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,GAAG,OAAO,CAAC,CAAC,IAAI,CAAC;IAC1D,OAAO,EAAE,MAAM,EAAE,YAAY,EAAE,WAAW,EAAE,OAAO,EAAE,CAAC;AACxD,CAAC"}

package/dist/core/linkGraph.d.ts

@@ -0,0 +1,19 @@
+import type Database from 'better-sqlite3';
+/**
+ * Post-crawl link-graph metrics, computed in-process from the `links` table (no new infra):
+ *  - **iPR** — internal PageRank (weighted: nav/footer links count ×0.2, body ×1), log-
+ *    normalised to 0–100. Surfaces where internal authority concentrates (research/14 §1).
+ *  - **click_depth** — shortest path from the homepage following BODY links only (the
+ *    "honest" depth; nav/footer would make everything depth-1). null = unreachable via body.
+ * Writes both back to pages.ipr / pages.click_depth for the crawl.
+ */
+/**
+ * Full post-crawl link-graph finalization for a crawl: in-degree (inlink_count) from the
+ * `links` table, then iPR + click_depth. Idempotent — safe to re-run. Used by the crawler at
+ * the end of a crawl AND by the audit as a self-heal when a crawl's link-graph was left
+ * unpopulated (e.g. the crawl process was killed before it could finalize). When seedKey is
+ * null the homepage is inferred as the highest-in-degree page (computed first, above).
+ */
+export declare function finalizeLinkGraph(db: Database.Database, crawlId: string, seedKey: string | null): void;
+export declare function computeLinkGraph(db: Database.Database, crawlId: string, seedKey: string | null): void;
+//# sourceMappingURL=linkGraph.d.ts.map

package/dist/core/linkGraph.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"linkGraph.d.ts","sourceRoot":"","sources":["../../src/core/linkGraph.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,QAAQ,MAAM,gBAAgB,CAAC;AAE3C;;;;;;;GAOG;AACH;;;;;;GAMG;AACH,wBAAgB,iBAAiB,CAAC,EAAE,EAAE,QAAQ,CAAC,QAAQ,EAAE,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,IAAI,GAAG,IAAI,CActG;AAED,wBAAgB,gBAAgB,CAAC,EAAE,EAAE,QAAQ,CAAC,QAAQ,EAAE,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,IAAI,GAAG,IAAI,CA2ErG"}

package/dist/core/linkGraph.js

@@ -0,0 +1,125 @@
+/**
+ * Post-crawl link-graph metrics, computed in-process from the `links` table (no new infra):
+ *  - **iPR** — internal PageRank (weighted: nav/footer links count ×0.2, body ×1), log-
+ *    normalised to 0–100. Surfaces where internal authority concentrates (research/14 §1).
+ *  - **click_depth** — shortest path from the homepage following BODY links only (the
+ *    "honest" depth; nav/footer would make everything depth-1). null = unreachable via body.
+ * Writes both back to pages.ipr / pages.click_depth for the crawl.
+ */
+/**
+ * Full post-crawl link-graph finalization for a crawl: in-degree (inlink_count) from the
+ * `links` table, then iPR + click_depth. Idempotent — safe to re-run. Used by the crawler at
+ * the end of a crawl AND by the audit as a self-heal when a crawl's link-graph was left
+ * unpopulated (e.g. the crawl process was killed before it could finalize). When seedKey is
+ * null the homepage is inferred as the highest-in-degree page (computed first, above).
+ */
+export function finalizeLinkGraph(db, crawlId, seedKey) {
+    db.prepare(`UPDATE pages SET inlink_count = (
+      SELECT COUNT(*) FROM links
+      WHERE links.target_key = pages.url_key AND links.is_internal = 1
+        AND links.source_key != pages.url_key AND links.crawl_id = pages.crawl_id
+    ) WHERE crawl_id = ?`).run(crawlId);
+    let seed = seedKey;
+    if (!seed) {
+        const r = db.prepare(`SELECT url_key FROM pages WHERE crawl_id=? ORDER BY inlink_count DESC LIMIT 1`).get(crawlId);
+        seed = r?.url_key ?? null;
+    }
+    computeLinkGraph(db, crawlId, seed);
+}
+export function computeLinkGraph(db, crawlId, seedKey) {
+    const nodes = db.prepare('SELECT url_key FROM pages WHERE crawl_id=?').all(crawlId).map(r => r.url_key);
+    const N = nodes.length;
+    if (N === 0)
+        return;
+    const idx = new Map(nodes.map((n, i) => [n, i]));
+    const edges = db.prepare('SELECT source_key s, target_key t, placement p FROM links WHERE crawl_id=? AND is_internal=1 AND source_key!=target_key').all(crawlId);
+    // ── iPR: weighted power iteration ──
+    const outW = new Float64Array(N);
+    const adj = Array.from({ length: N }, () => []);
+    for (const e of edges) {
+        const si = idx.get(e.s), ti = idx.get(e.t);
+        if (si === undefined || ti === undefined)
+            continue; // target not in this crawl
+        const w = e.p === 'body' ? 1 : 0.2;
+        adj[si].push({ t: ti, w });
+        outW[si] += w;
+    }
+    const d = 0.85;
+    let pr = new Float64Array(N).fill(1 / N);
+    for (let iter = 0; iter < 30; iter++) {
+        const next = new Float64Array(N).fill((1 - d) / N);
+        let dangling = 0;
+        for (let i = 0; i < N; i++)
+            if (outW[i] === 0)
+                dangling += pr[i];
+        const danglingShare = (d * dangling) / N;
+        for (let i = 0; i < N; i++) {
+            if (outW[i] === 0)
+                continue;
+            const share = (d * pr[i]) / outW[i];
+            for (const { t, w } of adj[i])
+                next[t] += share * w;
+        }
+        for (let i = 0; i < N; i++)
+            next[i] += danglingShare;
+        // L1-delta early exit — PageRank usually converges well before 30 iterations; stop once the
+        // vector barely moves so we don't spend the back half of the iterations for no change.
+        let delta = 0;
+        for (let i = 0; i < N; i++)
+            delta += Math.abs(next[i] - pr[i]);
+        pr = next;
+        if (delta < 1e-6)
+            break;
+    }
+    // log-normalise to 0–100 (pr values are all > 0 thanks to the (1-d)/N base)
+    let lo = Infinity, hi = -Infinity;
+    const logpr = new Float64Array(N);
+    for (let i = 0; i < N; i++) {
+        const l = Math.log(pr[i]);
+        logpr[i] = l;
+        if (l < lo)
+            lo = l;
+        if (l > hi)
+            hi = l;
+    }
+    const span = hi - lo || 1;
+    const ipr = new Float64Array(N);
+    for (let i = 0; i < N; i++)
+        ipr[i] = Math.round(((logpr[i] - lo) / span) * 100);
+    // ── click_depth: BFS from homepage over BODY links only ──
+    const bodyAdj = Array.from({ length: N }, () => []);
+    for (const e of edges) {
+        if (e.p !== 'body')
+            continue;
+        const si = idx.get(e.s), ti = idx.get(e.t);
+        if (si !== undefined && ti !== undefined)
+            bodyAdj[si].push(ti);
+    }
+    const depth = new Int32Array(N).fill(-1);
+    // BFS root: the seed's key, but if the bare homepage redirected (final URL keyed
+    // differently), fall back to the shallowest crawl-discovered page (depth 0 = seed).
+    let start = seedKey != null ? idx.get(seedKey) : undefined;
+    if (start === undefined) {
+        const root = db.prepare('SELECT url_key FROM pages WHERE crawl_id=? ORDER BY depth ASC, inlink_count DESC LIMIT 1').get(crawlId);
+        if (root)
+            start = idx.get(root.url_key);
+    }
+    if (start !== undefined) {
+        depth[start] = 0;
+        const q = [start];
+        for (let h = 0; h < q.length; h++) {
+            const u = q[h];
+            for (const v of bodyAdj[u])
+                if (depth[v] === -1) {
+                    depth[v] = depth[u] + 1;
+                    q.push(v);
+                }
+        }
+    }
+    const upd = db.prepare('UPDATE pages SET ipr=?, click_depth=? WHERE crawl_id=? AND url_key=?');
+    db.transaction(() => {
+        for (let i = 0; i < N; i++)
+            upd.run(ipr[i], depth[i] < 0 ? null : depth[i], crawlId, nodes[i]);
+    })();
+}
+//# sourceMappingURL=linkGraph.js.map

package/dist/core/linkGraph.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"linkGraph.js","sourceRoot":"","sources":["../../src/core/linkGraph.ts"],"names":[],"mappings":"AAEA;;;;;;;GAOG;AACH;;;;;;GAMG;AACH,MAAM,UAAU,iBAAiB,CAAC,EAAqB,EAAE,OAAe,EAAE,OAAsB;IAC9F,EAAE,CAAC,OAAO,CACR;;;;yBAIqB,CACtB,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;IACf,IAAI,IAAI,GAAG,OAAO,CAAC;IACnB,IAAI,CAAC,IAAI,EAAE,CAAC;QACV,MAAM,CAAC,GAAG,EAAE,CAAC,OAAO,CAAC,+EAA+E,CAAC,CAAC,GAAG,CAAC,OAAO,CAAoC,CAAC;QACtJ,IAAI,GAAG,CAAC,EAAE,OAAO,IAAI,IAAI,CAAC;IAC5B,CAAC;IACD,gBAAgB,CAAC,EAAE,EAAE,OAAO,EAAE,IAAI,CAAC,CAAC;AACtC,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,EAAqB,EAAE,OAAe,EAAE,OAAsB;IAC7F,MAAM,KAAK,GAAI,EAAE,CAAC,OAAO,CAAC,4CAA4C,CAAC,CAAC,GAAG,CAAC,OAAO,CAA2B,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC;IACnI,MAAM,CAAC,GAAG,KAAK,CAAC,MAAM,CAAC;IACvB,IAAI,CAAC,KAAK,CAAC;QAAE,OAAO;IACpB,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;IAEjD,MAAM,KAAK,GAAG,EAAE,CAAC,OAAO,CACtB,yHAAyH,CAC1H,CAAC,GAAG,CAAC,OAAO,CAA0C,CAAC;IAExD,sCAAsC;IACtC,MAAM,IAAI,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IACjC,MAAM,GAAG,GAAiC,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,GAAG,EAAE,CAAC,EAAE,CAAC,CAAC;IAC9E,KAAK,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC;QACtB,MAAM,EAAE,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,EAAE,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QAC3C,IAAI,EAAE,KAAK,SAAS,IAAI,EAAE,KAAK,SAAS;YAAE,SAAS,CAAC,2BAA2B;QAC/E,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,KAAK,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC;QACnC,GAAG,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,EAAE,EAAE,CAAC,EAAE,CAAC,CAAC;QAC3B,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,CAAC;IAChB,CAAC;IACD,MAAM,CAAC,GAAG,IAAI,CAAC;IACf,IAAI,EAAE,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IACzC,KAAK,IAAI,IAAI,GAAG,CAAC,EAAE,IAAI,GAAG,EAAE,EAAE,IAAI,EAAE,EAAE,CAAC;QACrC,MAAM,IAAI,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QACnD,IAAI,QAAQ,GAAG,CAAC,CAAC;QACjB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC;gBAAE,QAAQ,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;QACjE,MAAM,aAAa,GAAG,CAAC,CAAC,GAAG,QAAQ,CAAC,GAAG,CAAC,CAAC;QACzC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YAC3B,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC;gBAAE,SAAS;YAC5B,MAAM,KAAK,GAAG,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;YACpC,KAAK,MAAM,EAAE,CAAC,EAAE,CAAC,EAAE,IAAI,GAAG,CAAC,CAAC,CAAC;gBAAE,IAAI,CAAC,CAAC,CAAC,IAAI,KAAK,GAAG,CAAC,CAAC;QACtD,CAAC;QACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,IAAI,CAAC,CAAC,CAAC,IAAI,aAAa,CAAC;QACrD,4FAA4F;QAC5F,uFAAuF;QACvF,IAAI,KAAK,GAAG,CAAC,CAAC;QAAC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,KAAK,IAAI,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QAC9E,EAAE,GAAG,IAAI,CAAC;QACV,IAAI,KAAK,GAAG,IAAI;YAAE,MAAM;IAC1B,CAAC;IACD,4EAA4E;IAC5E,IAAI,EAAE,GAAG,QAAQ,EAAE,EAAE,GAAG,CAAC,QAAQ,CAAC;IAClC,MAAM,KAAK,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IAClC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAAC,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QAAC,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QAAC,IAAI,CAAC,GAAG,EAAE;YAAE,EAAE,GAAG,CAAC,CAAC;QAAC,IAAI,CAAC,GAAG,EAAE;YAAE,EAAE,GAAG,CAAC,CAAC;IAAC,CAAC;IAChH,MAAM,IAAI,GAAG,EAAE,GAAG,EAAE,IAAI,CAAC,CAAC;IAC1B,MAAM,GAAG,GAAG,IAAI,YAAY,CAAC,CAAC,CAAC,CAAC;IAChC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;QAAE,GAAG,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,EAAE,CAAC,GAAG,IAAI,CAAC,GAAG,GAAG,CAAC,CAAC;IAEhF,4DAA4D;IAC5D,MAAM,OAAO,GAAe,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,EAAE,GAAG,EAAE,CAAC,EAAE,CAAC,CAAC;IAChE,KAAK,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC;QACtB,IAAI,CAAC,CAAC,CAAC,KAAK,MAAM;YAAE,SAAS;QAC7B,MAAM,EAAE,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,EAAE,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QAC3C,IAAI,EAAE,KAAK,SAAS,IAAI,EAAE,KAAK,SAAS;YAAE,OAAO,CAAC,EAAE,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACjE,CAAC;IACD,MAAM,KAAK,GAAG,IAAI,UAAU,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACzC,iFAAiF;IACjF,oFAAoF;IACpF,IAAI,KAAK,GAAG,OAAO,IAAI,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;IAC3D,IAAI,KAAK,KAAK,SAAS,EAAE,CAAC;QACxB,MAAM,IAAI,GAAG,EAAE,CAAC,OAAO,CAAC,0FAA0F,CAAC,CAAC,GAAG,CAAC,OAAO,CAAoC,CAAC;QACpK,IAAI,IAAI;YAAE,KAAK,GAAG,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IAC1C,CAAC;IACD,IAAI,KAAK,KAAK,SAAS,EAAE,CAAC;QACxB,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;QACjB,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC;QAClB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YAClC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;YACf,KAAK,MAAM,CAAC,IAAI,OAAO,CAAC,CAAC,CAAC;gBAAE,IAAI,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC;oBAAC,KAAK,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;oBAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;gBAAC,CAAC;QAC1F,CAAC;IACH,CAAC;IAED,MAAM,GAAG,GAAG,EAAE,CAAC,OAAO,CAAC,sEAAsE,CAAC,CAAC;IAC/F,EAAE,CAAC,WAAW,CAAC,GAAG,EAAE;QAClB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE;YAAE,GAAG,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;IACjG,CAAC,CAAC,EAAE,CAAC;AACP,CAAC"}

package/dist/core/passageScore.d.ts

@@ -0,0 +1,19 @@
+/**
+ * For each ranking page (top GSC query, has body chunks), score every chunk against that query with
+ * the local cross-encoder and persist the single best-passage score — the proxy for how neural /
+ * AI search judges whether the page has a dense, extractable answer. On-demand (ML inference cost);
+ * bounded to pages that actually rank. Powers the `weak-passage-answer` check.
+ */
+export declare function scoreSitePassages(dataDir: string, siteUrl: string, opts?: {
+    limit?: number;
+    minImpressions?: number;
+}): Promise<{
+    scored: number;
+    flagged: number;
+    weakest: {
+        url: string;
+        query: string;
+        score: number;
+    }[];
+}>;
+//# sourceMappingURL=passageScore.d.ts.map

package/dist/core/passageScore.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"passageScore.d.ts","sourceRoot":"","sources":["../../src/core/passageScore.ts"],"names":[],"mappings":"AAKA;;;;;GAKG;AACH,wBAAsB,iBAAiB,CACrC,OAAO,EAAE,MAAM,EACf,OAAO,EAAE,MAAM,EACf,IAAI,GAAE;IAAE,KAAK,CAAC,EAAE,MAAM,CAAC;IAAC,cAAc,CAAC,EAAE,MAAM,CAAA;CAAO,GACrD,OAAO,CAAC;IAAE,MAAM,EAAE,MAAM,CAAC;IAAC,OAAO,EAAE,MAAM,CAAC;IAAC,OAAO,EAAE;QAAE,GAAG,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,EAAE,CAAA;CAAE,CAAC,CAwCxG"}

package/dist/core/passageScore.js

@@ -0,0 +1,59 @@
+import { AuditDatabase } from './AuditDatabase.js';
+import { dbPathFor } from './paths.js';
+import { gscFreshness } from './gscFreshness.js';
+import { maxPassageScore } from './reranker.js';
+/**
+ * For each ranking page (top GSC query, has body chunks), score every chunk against that query with
+ * the local cross-encoder and persist the single best-passage score — the proxy for how neural /
+ * AI search judges whether the page has a dense, extractable answer. On-demand (ML inference cost);
+ * bounded to pages that actually rank. Powers the `weak-passage-answer` check.
+ */
+export async function scoreSitePassages(dataDir, siteUrl, opts = {}) {
+    const db = new AuditDatabase(dbPathFor(dataDir, siteUrl));
+    try {
+        const maxDate = gscFreshness(db.db).effectiveMax;
+        if (!maxDate)
+            return { scored: 0, flagged: 0, weakest: [] };
+        const minImpr = opts.minImpressions ?? 50;
+        const win = `date > date('${maxDate}', '-28 days') AND date <= '${maxDate}'`;
+        const pages = db.db.prepare(`SELECT s.page_key urlKey, s.query query, s.impr impressions, p.body_chunks bc FROM
+        (SELECT page_key, query, SUM(impressions) impr, ROW_NUMBER() OVER (PARTITION BY page_key ORDER BY SUM(impressions) DESC) rn
+         FROM search_analytics WHERE query IS NOT NULL AND page_key IS NOT NULL AND ${win} GROUP BY page_key, query) s
+       JOIN pages p ON p.url_key = s.page_key
+       WHERE s.rn = 1 AND p.indexable = 1 AND p.body_chunks IS NOT NULL AND s.impr >= ${minImpr}
+       ORDER BY s.impr DESC LIMIT ${opts.limit ? Math.floor(opts.limit) : 500}`).all();
+        const upd = db.db.prepare(`UPDATE pages SET max_passage_score=@s, max_passage_query=@q, max_passage_impr=@i, max_passage_at=datetime('now') WHERE url_key=@u`);
+        let scored = 0, flagged = 0;
+        const weakest = [];
+        for (const pg of pages) {
+            let chunks;
+            try {
+                chunks = JSON.parse(pg.bc);
+            }
+            catch {
+                continue;
+            }
+            const passages = chunks
+                .map(c => `${c.heading ? `${c.heading}. ` : ''}${c.text || ''}`.trim())
+                .filter(t => t.length > 20);
+            if (!passages.length)
+                continue;
+            const { max } = await maxPassageScore(pg.query, passages);
+            if (!Number.isFinite(max))
+                continue; // empty/whitespace query → no scores; don't persist -Infinity
+            const score = Math.round(max * 100) / 100;
+            upd.run({ s: score, q: pg.query, i: pg.impressions, u: pg.urlKey });
+            scored++;
+            if (score < 3) {
+                flagged++;
+                weakest.push({ url: pg.urlKey, query: pg.query, score });
+            }
+        }
+        weakest.sort((a, b) => a.score - b.score);
+        return { scored, flagged, weakest: weakest.slice(0, 20) };
+    }
+    finally {
+        db.close();
+    }
+}
+//# sourceMappingURL=passageScore.js.map

package/dist/core/passageScore.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"passageScore.js","sourceRoot":"","sources":["../../src/core/passageScore.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,MAAM,oBAAoB,CAAC;AACnD,OAAO,EAAE,SAAS,EAAE,MAAM,YAAY,CAAC;AACvC,OAAO,EAAE,YAAY,EAAE,MAAM,mBAAmB,CAAC;AACjD,OAAO,EAAE,eAAe,EAAE,MAAM,eAAe,CAAC;AAEhD;;;;;GAKG;AACH,MAAM,CAAC,KAAK,UAAU,iBAAiB,CACrC,OAAe,EACf,OAAe,EACf,OAAoD,EAAE;IAEtD,MAAM,EAAE,GAAG,IAAI,aAAa,CAAC,SAAS,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC,CAAC;IAC1D,IAAI,CAAC;QACH,MAAM,OAAO,GAAG,YAAY,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,YAAY,CAAC;QACjD,IAAI,CAAC,OAAO;YAAE,OAAO,EAAE,MAAM,EAAE,CAAC,EAAE,OAAO,EAAE,CAAC,EAAE,OAAO,EAAE,EAAE,EAAE,CAAC;QAC5D,MAAM,OAAO,GAAG,IAAI,CAAC,cAAc,IAAI,EAAE,CAAC;QAC1C,MAAM,GAAG,GAAG,gBAAgB,OAAO,+BAA+B,OAAO,GAAG,CAAC;QAC7E,MAAM,KAAK,GAAG,EAAE,CAAC,EAAE,CAAC,OAAO,CACzB;;sFAEgF,GAAG;;wFAED,OAAO;oCAC3D,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,EAAE,CAC1E,CAAC,GAAG,EAA0E,CAAC;QAEhF,MAAM,GAAG,GAAG,EAAE,CAAC,EAAE,CAAC,OAAO,CACvB,mIAAmI,CACpI,CAAC;QACF,IAAI,MAAM,GAAG,CAAC,EAAE,OAAO,GAAG,CAAC,CAAC;QAC5B,MAAM,OAAO,GAAoD,EAAE,CAAC;QACpE,KAAK,MAAM,EAAE,IAAI,KAAK,EAAE,CAAC;YACvB,IAAI,MAA6C,CAAC;YAClD,IAAI,CAAC;gBAAC,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;YAAC,CAAC;YAAC,MAAM,CAAC;gBAAC,SAAS;YAAC,CAAC;YACvD,MAAM,QAAQ,GAAG,MAAM;iBACpB,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,IAAI,IAAI,EAAE,EAAE,CAAC,IAAI,EAAE,CAAC;iBACtE,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,EAAE,CAAC,CAAC;YAC9B,IAAI,CAAC,QAAQ,CAAC,MAAM;gBAAE,SAAS;YAC/B,MAAM,EAAE,GAAG,EAAE,GAAG,MAAM,eAAe,CAAC,EAAE,CAAC,KAAK,EAAE,QAAQ,CAAC,CAAC;YAC1D,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC;gBAAE,SAAS,CAAC,8DAA8D;YACnG,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,GAAG,GAAG,CAAC,GAAG,GAAG,CAAC;YAC1C,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,EAAE,CAAC,KAAK,EAAE,CAAC,EAAE,EAAE,CAAC,WAAW,EAAE,CAAC,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,CAAC;YACpE,MAAM,EAAE,CAAC;YACT,IAAI,KAAK,GAAG,CAAC,EAAE,CAAC;gBAAC,OAAO,EAAE,CAAC;gBAAC,OAAO,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,EAAE,CAAC,MAAM,EAAE,KAAK,EAAE,EAAE,CAAC,KAAK,EAAE,KAAK,EAAE,CAAC,CAAC;YAAC,CAAC;QACzF,CAAC;QACD,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;QAC1C,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAC;IAC5D,CAAC;YAAS,CAAC;QACT,EAAE,CAAC,KAAK,EAAE,CAAC;IACb,CAAC;AACH,CAAC"}

package/dist/core/paths.d.ts

@@ -0,0 +1,5 @@
+/** Turn a GSC property identifier into a safe filename stem. */
+export declare function sanitizeProperty(siteUrl: string): string;
+/** Absolute path to a property's SQLite database (one DB per property). */
+export declare function dbPathFor(dataDir: string, siteUrl: string): string;
+//# sourceMappingURL=paths.d.ts.map

package/dist/core/paths.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"paths.d.ts","sourceRoot":"","sources":["../../src/core/paths.ts"],"names":[],"mappings":"AAEA,gEAAgE;AAChE,wBAAgB,gBAAgB,CAAC,OAAO,EAAE,MAAM,GAAG,MAAM,CAOxD;AAED,2EAA2E;AAC3E,wBAAgB,SAAS,CAAC,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,MAAM,CAElE"}

package/dist/core/paths.js

@@ -0,0 +1,15 @@
+import path from 'node:path';
+/** Turn a GSC property identifier into a safe filename stem. */
+export function sanitizeProperty(siteUrl) {
+    return siteUrl
+        .replace(/^sc-domain:/, '')
+        .replace(/^https?:\/\//, '')
+        .replace(/[^a-zA-Z0-9.-]/g, '_')
+        .replace(/_+/g, '_')
+        .replace(/^_|_$/g, '');
+}
+/** Absolute path to a property's SQLite database (one DB per property). */
+export function dbPathFor(dataDir, siteUrl) {
+    return path.join(dataDir, `${sanitizeProperty(siteUrl)}.db`);
+}
+//# sourceMappingURL=paths.js.map

package/dist/core/paths.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"paths.js","sourceRoot":"","sources":["../../src/core/paths.ts"],"names":[],"mappings":"AAAA,OAAO,IAAI,MAAM,WAAW,CAAC;AAE7B,gEAAgE;AAChE,MAAM,UAAU,gBAAgB,CAAC,OAAe;IAC9C,OAAO,OAAO;SACX,OAAO,CAAC,aAAa,EAAE,EAAE,CAAC;SAC1B,OAAO,CAAC,cAAc,EAAE,EAAE,CAAC;SAC3B,OAAO,CAAC,iBAAiB,EAAE,GAAG,CAAC;SAC/B,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC;SACnB,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,CAAC;AAC3B,CAAC;AAED,2EAA2E;AAC3E,MAAM,UAAU,SAAS,CAAC,OAAe,EAAE,OAAe;IACxD,OAAO,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,GAAG,gBAAgB,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;AAC/D,CAAC"}