@houtini/seo-audit-console 0.2.0

package/dist/core/queryData.d.ts

@@ -0,0 +1,35 @@
+export declare const QUERYABLE_TABLES: readonly ["pages", "links", "search_analytics", "url_inspection", "sitemap_urls", "findings", "image_assets", "page_backlinks"];
+export type QueryTable = (typeof QUERYABLE_TABLES)[number];
+export declare const FILTER_OPS: readonly ["eq", "ne", "gt", "gte", "lt", "lte", "like", "in"];
+export type FilterOp = (typeof FILTER_OPS)[number];
+export declare const METRIC_FNS: readonly ["sum", "avg", "min", "max"];
+export type MetricFn = (typeof METRIC_FNS)[number];
+export interface QueryFilter {
+    column: string;
+    op: FilterOp;
+    value: unknown;
+}
+export interface QueryMetric {
+    fn: MetricFn;
+    column: string;
+}
+export interface QueryDataInput {
+    table: QueryTable;
+    mode?: 'aggregate' | 'rows';
+    groupBy?: string[];
+    metrics?: QueryMetric[];
+    filters?: QueryFilter[];
+    orderBy?: string;
+    orderDir?: 'asc' | 'desc';
+    limit?: number;
+    offset?: number;
+    columns?: string[];
+}
+export interface QueryDataResult {
+    markdown: string;
+    structured: Record<string, unknown>;
+}
+/** Truncate a cell value at CELL_CAP chars — always loud (trailing …). */
+export declare function truncateCell(v: unknown): unknown;
+export declare function runQueryData(dbPath: string, input: QueryDataInput): QueryDataResult;
+//# sourceMappingURL=queryData.d.ts.map

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

@@ -0,0 +1 @@
+{"version":3,"file":"queryData.d.ts","sourceRoot":"","sources":["../../src/core/queryData.ts"],"names":[],"mappings":"AASA,eAAO,MAAM,gBAAgB,iIAGnB,CAAC;AACX,MAAM,MAAM,UAAU,GAAG,CAAC,OAAO,gBAAgB,CAAC,CAAC,MAAM,CAAC,CAAC;AAE3D,eAAO,MAAM,UAAU,+DAAgE,CAAC;AACxF,MAAM,MAAM,QAAQ,GAAG,CAAC,OAAO,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC;AAKnD,eAAO,MAAM,UAAU,uCAAwC,CAAC;AAChE,MAAM,MAAM,QAAQ,GAAG,CAAC,OAAO,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC;AAEnD,MAAM,WAAW,WAAW;IAAG,MAAM,EAAE,MAAM,CAAC;IAAC,EAAE,EAAE,QAAQ,CAAC;IAAC,KAAK,EAAE,OAAO,CAAA;CAAE;AAC7E,MAAM,WAAW,WAAW;IAAG,EAAE,EAAE,QAAQ,CAAC;IAAC,MAAM,EAAE,MAAM,CAAA;CAAE;AAE7D,MAAM,WAAW,cAAc;IAC7B,KAAK,EAAE,UAAU,CAAC;IAClB,IAAI,CAAC,EAAE,WAAW,GAAG,MAAM,CAAC;IAC5B,OAAO,CAAC,EAAE,MAAM,EAAE,CAAC;IACnB,OAAO,CAAC,EAAE,WAAW,EAAE,CAAC;IACxB,OAAO,CAAC,EAAE,WAAW,EAAE,CAAC;IACxB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,KAAK,GAAG,MAAM,CAAC;IAC1B,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,MAAM,EAAE,CAAC;CACpB;AAED,MAAM,WAAW,eAAe;IAC9B,QAAQ,EAAE,MAAM,CAAC;IACjB,UAAU,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACrC;AAiBD,0EAA0E;AAC1E,wBAAgB,YAAY,CAAC,CAAC,EAAE,OAAO,GAAG,OAAO,CAGhD;AAWD,wBAAgB,YAAY,CAAC,MAAM,EAAE,MAAM,EAAE,KAAK,EAAE,cAAc,GAAG,eAAe,CAenF"}

package/dist/core/queryData.js

@@ -0,0 +1,200 @@
+/**
+ * queryData — the composition-layer multiplier: aggregate IN the database, return
+ * answers, not rows. Read-only (SQLite opened readonly, SELECTs only). Every
+ * identifier (table, column) is validated against the table's actual columns via
+ * PRAGMA table_info; every VALUE travels as a bound parameter — nothing user-supplied
+ * is ever interpolated into SQL.
+ */
+import Database from 'better-sqlite3';
+export const QUERYABLE_TABLES = [
+    'pages', 'links', 'search_analytics', 'url_inspection', 'sitemap_urls',
+    'findings', 'image_assets', 'page_backlinks',
+];
+export const FILTER_OPS = ['eq', 'ne', 'gt', 'gte', 'lt', 'lte', 'like', 'in'];
+const OP_SQL = {
+    eq: '=', ne: '<>', gt: '>', gte: '>=', lt: '<', lte: '<=', like: 'LIKE',
+};
+export const METRIC_FNS = ['sum', 'avg', 'min', 'max'];
+/** Sensible default columns per table for rows mode — the fat columns (body_chunks,
+ *  json_ld, evidence…) stay out unless explicitly requested. */
+const DEFAULT_ROW_COLUMNS = {
+    pages: ['url', 'status_code', 'indexable', 'title', 'h1', 'word_count', 'ipr', 'click_depth', 'inlink_count', 'canonical_key'],
+    links: ['source_key', 'target_key', 'anchor_text', 'is_internal', 'placement', 'rel'],
+    search_analytics: ['date', 'query', 'page_key', 'clicks', 'impressions', 'ctr', 'position'],
+    url_inspection: ['url_key', 'verdict', 'coverage_state', 'google_canonical', 'user_canonical', 'last_crawl_time'],
+    sitemap_urls: ['url', 'lastmod'],
+    findings: ['id', 'run_id', 'check_id', 'category', 'severity', 'url_key', 'priority'],
+    image_assets: ['url', 'bytes', 'status', 'content_type', 'used_on'],
+    page_backlinks: ['url', 'backlinks', 'referring_domains', 'dofollow', 'status_code'],
+};
+const CELL_CAP = 120;
+/** Truncate a cell value at CELL_CAP chars — always loud (trailing …). */
+export function truncateCell(v) {
+    if (typeof v !== 'string' || v.length <= CELL_CAP)
+        return v;
+    return v.slice(0, CELL_CAP) + '…';
+}
+const mdCell = (v) => {
+    if (v == null)
+        return '—';
+    return String(truncateCell(v)).replace(/\|/g, '\\|').replace(/\r?\n/g, ' ');
+};
+const quoteIdent = (name) => `"${name.replace(/"/g, '""')}"`;
+const NUMERIC_AFFINITY = /INT|REAL|FLOA|DOUB|NUM|DECI/i;
+export function runQueryData(dbPath, input) {
+    let db;
+    try {
+        db = new Database(dbPath, { readonly: true, fileMustExist: true });
+    }
+    catch {
+        return {
+            markdown: `No database at ${dbPath} — run refresh_property (or start_crawl / sync_gsc) first.`,
+            structured: { error: 'no_database', dbPath },
+        };
+    }
+    try {
+        return execute(db, input);
+    }
+    finally {
+        db.close();
+    }
+}
+function execute(db, input) {
+    const table = input.table;
+    if (!QUERYABLE_TABLES.includes(table))
+        throw new Error(`Unknown table: ${table}`);
+    // The column whitelist IS the table's real columns (PRAGMA table_info) — anything
+    // not in it is hard-rejected. Numeric columns (INTEGER/REAL affinity) additionally
+    // qualify for sum/avg/min/max.
+    const info = db.prepare(`PRAGMA table_info(${quoteIdent(table)})`).all();
+    if (!info.length) {
+        return { markdown: `Table ${table} has no columns in this database (nothing synced yet?).`, structured: { error: 'empty_table_schema', table } };
+    }
+    const allCols = new Set(info.map(c => c.name));
+    const numericCols = new Set(info.filter(c => NUMERIC_AFFINITY.test(c.type ?? '')).map(c => c.name));
+    const requireCol = (name, context) => {
+        if (!allCols.has(name))
+            throw new Error(`Unknown column in ${context}: "${name}". Columns on ${table}: ${[...allCols].join(', ')}`);
+        return name;
+    };
+    const requireNumeric = (name, context) => {
+        requireCol(name, context);
+        if (!numericCols.has(name))
+            throw new Error(`Column "${name}" is not numeric — ${context} needs one of: ${[...numericCols].join(', ')}`);
+        return name;
+    };
+    // WHERE — column names from the whitelist, values ALWAYS as bound parameters.
+    const whereParts = [];
+    const params = [];
+    for (const f of input.filters ?? []) {
+        const col = quoteIdent(requireCol(f.column, 'filters'));
+        if (f.op === 'in') {
+            if (!Array.isArray(f.value) || !f.value.length)
+                throw new Error(`Filter op "in" on ${f.column} needs a non-empty array value.`);
+            if (f.value.length > 100)
+                throw new Error(`Filter op "in" on ${f.column}: max 100 values.`);
+            whereParts.push(`${col} IN (${f.value.map(() => '?').join(',')})`);
+            params.push(...f.value);
+        }
+        else if (f.value === null && (f.op === 'eq' || f.op === 'ne')) {
+            whereParts.push(`${col} IS ${f.op === 'eq' ? '' : 'NOT '}NULL`);
+        }
+        else {
+            if (f.value === null || Array.isArray(f.value) || typeof f.value === 'object') {
+                throw new Error(`Filter op "${f.op}" on ${f.column} needs a scalar value.`);
+            }
+            whereParts.push(`${col} ${OP_SQL[f.op]} ?`);
+            params.push(f.value);
+        }
+    }
+    const where = whereParts.length ? ` WHERE ${whereParts.join(' AND ')}` : '';
+    const from = `FROM ${quoteIdent(table)}${where}`;
+    const mode = input.mode ?? 'aggregate';
+    const orderDir = input.orderDir === 'asc' ? 'ASC' : 'DESC';
+    // Total matching rows — both modes report it (the honest denominator).
+    const rowsTotal = db.prepare(`SELECT COUNT(*) n ${from}`).get(...params).n;
+    if (mode === 'aggregate') {
+        const groupBy = (input.groupBy ?? []).map(c => requireCol(c, 'groupBy'));
+        const metrics = (input.metrics ?? []).map(m => {
+            if (!METRIC_FNS.includes(m.fn))
+                throw new Error(`Unknown metric fn: ${m.fn}`);
+            return { fn: m.fn, column: requireNumeric(m.column, `metrics (${m.fn})`), alias: `${m.fn}_${m.column}` };
+        });
+        const sel = [
+            ...groupBy.map(quoteIdent),
+            'COUNT(*) AS count',
+            ...metrics.map(m => `${m.fn.toUpperCase()}(${quoteIdent(m.column)}) AS ${quoteIdent(m.alias)}`),
+        ].join(', ');
+        const groupClause = groupBy.length ? ` GROUP BY ${groupBy.map(quoteIdent).join(', ')}` : '';
+        // orderBy: a groupBy column, 'count', or a requested metric alias (e.g. sum_clicks).
+        const orderable = new Set(['count', ...groupBy, ...metrics.map(m => m.alias)]);
+        const orderCol = input.orderBy ?? 'count';
+        if (!orderable.has(orderCol))
+            throw new Error(`orderBy "${orderCol}" must be one of: ${[...orderable].join(', ')}`);
+        const limit = Math.min(Math.max(input.limit ?? 25, 1), 500);
+        const sql = `SELECT ${sel} ${from}${groupClause} ORDER BY ${quoteIdent(orderCol)} ${orderDir} LIMIT ${limit}`;
+        const groups = db.prepare(sql).all(...params);
+        const groupsTotal = groupBy.length
+            ? db.prepare(`SELECT COUNT(*) n FROM (SELECT 1 ${from}${groupClause})`).get(...params).n
+            : 1;
+        // Round avg to 2dp; attach SF-style percentages.
+        const out = groups.map(g => {
+            const row = {};
+            for (const c of groupBy)
+                row[c] = truncateCell(g[c]);
+            const count = Number(g.count) || 0;
+            row.count = count;
+            row.pct = rowsTotal > 0 ? Math.round((count / rowsTotal) * 1000) / 10 : 0;
+            for (const m of metrics) {
+                const v = g[m.alias];
+                row[m.alias] = m.fn === 'avg' && typeof v === 'number' ? Math.round(v * 100) / 100 : v;
+            }
+            return row;
+        });
+        const headerCols = [...groupBy, 'count', '%', ...metrics.map(m => m.alias)];
+        const header = `| ${headerCols.join(' | ')} |\n|${headerCols.map(() => '---').join('|')}|`;
+        const body = out.map(r => `| ${[...groupBy.map(c => mdCell(r[c])), String(r.count), `${r.pct}%`, ...metrics.map(m => mdCell(r[m.alias]))].join(' | ')} |`).join('\n');
+        const shownPct = rowsTotal > 0
+            ? Math.round((out.reduce((s, r) => s + r.count, 0) / rowsTotal) * 1000) / 10
+            : 0;
+        const totalLine = groupBy.length
+            ? `Total: ${rowsTotal.toLocaleString('en-US')} rows across ${groupsTotal.toLocaleString('en-US')} groups — showing ${out.length}${groupsTotal > out.length ? ` (covering ${shownPct}% of rows; raise limit or add filters for the rest)` : ''}.`
+            : `Total: ${rowsTotal.toLocaleString('en-US')} rows (no groupBy — single aggregate).`;
+        const markdown = out.length ? `${header}\n${body}\n\n${totalLine}` : `No rows match. ${totalLine}`;
+        return {
+            markdown,
+            structured: { table, mode, groupBy, metrics: metrics.map(({ fn, column }) => ({ fn, column })), groups: out, groupsTotal, groupsReturned: out.length, rowsTotal },
+        };
+    }
+    // rows mode — token discipline: curated default columns, 120-char cells, loud paging footer.
+    const cols = (input.columns?.length ? input.columns : DEFAULT_ROW_COLUMNS[table]).map(c => requireCol(c, 'columns'));
+    const orderCol = input.orderBy ? requireCol(input.orderBy, 'orderBy') : null;
+    const limit = Math.min(Math.max(input.limit ?? 50, 1), 500);
+    const offset = Math.max(input.offset ?? 0, 0);
+    const sql = `SELECT ${cols.map(quoteIdent).join(', ')} ${from}${orderCol ? ` ORDER BY ${quoteIdent(orderCol)} ${orderDir}` : ''} LIMIT ${limit} OFFSET ${offset}`;
+    const raw = db.prepare(sql).all(...params);
+    let truncated = false;
+    const rows = raw.map(r => {
+        const o = {};
+        for (const c of cols) {
+            const t = truncateCell(r[c]);
+            if (t !== r[c])
+                truncated = true;
+            o[c] = t;
+        }
+        return o;
+    });
+    const first = rows.length ? offset + 1 : 0;
+    const last = offset + rows.length;
+    const nextOffset = last < rowsTotal ? last : null;
+    const footer = `Showing ${first}–${last} of ${rowsTotal.toLocaleString('en-US')}${nextOffset != null ? `; next offset ${nextOffset}` : ' (end)'}.` +
+        (truncated ? ` Cells over ${CELL_CAP} chars truncated (…) — use mode:aggregate or fewer columns for full values via SQL.` : '');
+    const header = `| ${cols.join(' | ')} |\n|${cols.map(() => '---').join('|')}|`;
+    const body = rows.map(r => `| ${cols.map(c => mdCell(r[c])).join(' | ')} |`).join('\n');
+    const markdown = rows.length ? `${header}\n${body}\n\n${footer}` : `No rows match (${rowsTotal} total in ${table} for these filters).`;
+    return {
+        markdown,
+        structured: { table, mode, columns: cols, rows, rowsReturned: rows.length, rowsTotal, offset, nextOffset, cellsTruncated: truncated },
+    };
+}
+//# sourceMappingURL=queryData.js.map

package/dist/core/queryData.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"queryData.js","sourceRoot":"","sources":["../../src/core/queryData.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,OAAO,QAAQ,MAAM,gBAAgB,CAAC;AAEtC,MAAM,CAAC,MAAM,gBAAgB,GAAG;IAC9B,OAAO,EAAE,OAAO,EAAE,kBAAkB,EAAE,gBAAgB,EAAE,cAAc;IACtE,UAAU,EAAE,cAAc,EAAE,gBAAgB;CACpC,CAAC;AAGX,MAAM,CAAC,MAAM,UAAU,GAAG,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,CAAU,CAAC;AAExF,MAAM,MAAM,GAA4C;IACtD,EAAE,EAAE,GAAG,EAAE,EAAE,EAAE,IAAI,EAAE,EAAE,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,EAAE,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM;CACxE,CAAC;AAEF,MAAM,CAAC,MAAM,UAAU,GAAG,CAAC,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,CAAU,CAAC;AAwBhE;gEACgE;AAChE,MAAM,mBAAmB,GAAiC;IACxD,KAAK,EAAE,CAAC,KAAK,EAAE,aAAa,EAAE,WAAW,EAAE,OAAO,EAAE,IAAI,EAAE,YAAY,EAAE,KAAK,EAAE,aAAa,EAAE,cAAc,EAAE,eAAe,CAAC;IAC9H,KAAK,EAAE,CAAC,YAAY,EAAE,YAAY,EAAE,aAAa,EAAE,aAAa,EAAE,WAAW,EAAE,KAAK,CAAC;IACrF,gBAAgB,EAAE,CAAC,MAAM,EAAE,OAAO,EAAE,UAAU,EAAE,QAAQ,EAAE,aAAa,EAAE,KAAK,EAAE,UAAU,CAAC;IAC3F,cAAc,EAAE,CAAC,SAAS,EAAE,SAAS,EAAE,gBAAgB,EAAE,kBAAkB,EAAE,gBAAgB,EAAE,iBAAiB,CAAC;IACjH,YAAY,EAAE,CAAC,KAAK,EAAE,SAAS,CAAC;IAChC,QAAQ,EAAE,CAAC,IAAI,EAAE,QAAQ,EAAE,UAAU,EAAE,UAAU,EAAE,UAAU,EAAE,SAAS,EAAE,UAAU,CAAC;IACrF,YAAY,EAAE,CAAC,KAAK,EAAE,OAAO,EAAE,QAAQ,EAAE,cAAc,EAAE,SAAS,CAAC;IACnE,cAAc,EAAE,CAAC,KAAK,EAAE,WAAW,EAAE,mBAAmB,EAAE,UAAU,EAAE,aAAa,CAAC;CACrF,CAAC;AAEF,MAAM,QAAQ,GAAG,GAAG,CAAC;AAErB,0EAA0E;AAC1E,MAAM,UAAU,YAAY,CAAC,CAAU;IACrC,IAAI,OAAO,CAAC,KAAK,QAAQ,IAAI,CAAC,CAAC,MAAM,IAAI,QAAQ;QAAE,OAAO,CAAC,CAAC;IAC5D,OAAO,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,QAAQ,CAAC,GAAG,GAAG,CAAC;AACpC,CAAC;AAED,MAAM,MAAM,GAAG,CAAC,CAAU,EAAU,EAAE;IACpC,IAAI,CAAC,IAAI,IAAI;QAAE,OAAO,GAAG,CAAC;IAC1B,OAAO,MAAM,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;AAC9E,CAAC,CAAC;AAEF,MAAM,UAAU,GAAG,CAAC,IAAY,EAAU,EAAE,CAAC,IAAI,IAAI,CAAC,OAAO,CAAC,IAAI,EAAE,IAAI,CAAC,GAAG,CAAC;AAE7E,MAAM,gBAAgB,GAAG,8BAA8B,CAAC;AAExD,MAAM,UAAU,YAAY,CAAC,MAAc,EAAE,KAAqB;IAChE,IAAI,EAAqB,CAAC;IAC1B,IAAI,CAAC;QACH,EAAE,GAAG,IAAI,QAAQ,CAAC,MAAM,EAAE,EAAE,QAAQ,EAAE,IAAI,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;IACrE,CAAC;IAAC,MAAM,CAAC;QACP,OAAO;YACL,QAAQ,EAAE,kBAAkB,MAAM,4DAA4D;YAC9F,UAAU,EAAE,EAAE,KAAK,EAAE,aAAa,EAAE,MAAM,EAAE;SAC7C,CAAC;IACJ,CAAC;IACD,IAAI,CAAC;QACH,OAAO,OAAO,CAAC,EAAE,EAAE,KAAK,CAAC,CAAC;IAC5B,CAAC;YAAS,CAAC;QACT,EAAE,CAAC,KAAK,EAAE,CAAC;IACb,CAAC;AACH,CAAC;AAED,SAAS,OAAO,CAAC,EAAqB,EAAE,KAAqB;IAC3D,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC;IAC1B,IAAI,CAAC,gBAAgB,CAAC,QAAQ,CAAC,KAAK,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,kBAAkB,KAAK,EAAE,CAAC,CAAC;IAElF,kFAAkF;IAClF,mFAAmF;IACnF,+BAA+B;IAC/B,MAAM,IAAI,GAAG,EAAE,CAAC,OAAO,CAAC,qBAAqB,UAAU,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,EAAsC,CAAC;IAC7G,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC;QACjB,OAAO,EAAE,QAAQ,EAAE,SAAS,KAAK,yDAAyD,EAAE,UAAU,EAAE,EAAE,KAAK,EAAE,oBAAoB,EAAE,KAAK,EAAE,EAAE,CAAC;IACnJ,CAAC;IACD,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC;IAC/C,MAAM,WAAW,GAAG,IAAI,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,gBAAgB,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC;IACpG,MAAM,UAAU,GAAG,CAAC,IAAY,EAAE,OAAe,EAAU,EAAE;QAC3D,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,IAAI,CAAC;YAAE,MAAM,IAAI,KAAK,CAAC,qBAAqB,OAAO,MAAM,IAAI,iBAAiB,KAAK,KAAK,CAAC,GAAG,OAAO,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACpI,OAAO,IAAI,CAAC;IACd,CAAC,CAAC;IACF,MAAM,cAAc,GAAG,CAAC,IAAY,EAAE,OAAe,EAAU,EAAE;QAC/D,UAAU,CAAC,IAAI,EAAE,OAAO,CAAC,CAAC;QAC1B,IAAI,CAAC,WAAW,CAAC,GAAG,CAAC,IAAI,CAAC;YAAE,MAAM,IAAI,KAAK,CAAC,WAAW,IAAI,sBAAsB,OAAO,kBAAkB,CAAC,GAAG,WAAW,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACzI,OAAO,IAAI,CAAC;IACd,CAAC,CAAC;IAEF,8EAA8E;IAC9E,MAAM,UAAU,GAAa,EAAE,CAAC;IAChC,MAAM,MAAM,GAAc,EAAE,CAAC;IAC7B,KAAK,MAAM,CAAC,IAAI,KAAK,CAAC,OAAO,IAAI,EAAE,EAAE,CAAC;QACpC,MAAM,GAAG,GAAG,UAAU,CAAC,UAAU,CAAC,CAAC,CAAC,MAAM,EAAE,SAAS,CAAC,CAAC,CAAC;QACxD,IAAI,CAAC,CAAC,EAAE,KAAK,IAAI,EAAE,CAAC;YAClB,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,MAAM;gBAAE,MAAM,IAAI,KAAK,CAAC,qBAAqB,CAAC,CAAC,MAAM,iCAAiC,CAAC,CAAC;YAChI,IAAI,CAAC,CAAC,KAAK,CAAC,MAAM,GAAG,GAAG;gBAAE,MAAM,IAAI,KAAK,CAAC,qBAAqB,CAAC,CAAC,MAAM,mBAAmB,CAAC,CAAC;YAC5F,UAAU,CAAC,IAAI,CAAC,GAAG,GAAG,QAAQ,CAAC,CAAC,KAAK,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;YACnE,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;QAC1B,CAAC;aAAM,IAAI,CAAC,CAAC,KAAK,KAAK,IAAI,IAAI,CAAC,CAAC,CAAC,EAAE,KAAK,IAAI,IAAI,CAAC,CAAC,EAAE,KAAK,IAAI,CAAC,EAAE,CAAC;YAChE,UAAU,CAAC,IAAI,CAAC,GAAG,GAAG,OAAO,CAAC,CAAC,EAAE,KAAK,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,MAAM,CAAC,CAAC;QAClE,CAAC;aAAM,CAAC;YACN,IAAI,CAAC,CAAC,KAAK,KAAK,IAAI,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,OAAO,CAAC,CAAC,KAAK,KAAK,QAAQ,EAAE,CAAC;gBAC9E,MAAM,IAAI,KAAK,CAAC,cAAc,CAAC,CAAC,EAAE,QAAQ,CAAC,CAAC,MAAM,wBAAwB,CAAC,CAAC;YAC9E,CAAC;YACD,UAAU,CAAC,IAAI,CAAC,GAAG,GAAG,IAAI,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,IAAI,CAAC,CAAC;YAC5C,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC;QACvB,CAAC;IACH,CAAC;IACD,MAAM,KAAK,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC,CAAC,UAAU,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;IAC5E,MAAM,IAAI,GAAG,QAAQ,UAAU,CAAC,KAAK,CAAC,GAAG,KAAK,EAAE,CAAC;IAEjD,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,IAAI,WAAW,CAAC;IACvC,MAAM,QAAQ,GAAG,KAAK,CAAC,QAAQ,KAAK,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC;IAE3D,uEAAuE;IACvE,MAAM,SAAS,GAAI,EAAE,CAAC,OAAO,CAAC,qBAAqB,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,GAAG,MAAM,CAAmB,CAAC,CAAC,CAAC;IAE9F,IAAI,IAAI,KAAK,WAAW,EAAE,CAAC;QACzB,MAAM,OAAO,GAAG,CAAC,KAAK,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,UAAU,CAAC,CAAC,EAAE,SAAS,CAAC,CAAC,CAAC;QACzE,MAAM,OAAO,GAAG,CAAC,KAAK,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE;YAC5C,IAAI,CAAC,UAAU,CAAC,QAAQ,CAAC,CAAC,CAAC,EAAE,CAAC;gBAAE,MAAM,IAAI,KAAK,CAAC,sBAAsB,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;YAC9E,OAAO,EAAE,EAAE,EAAE,CAAC,CAAC,EAAE,EAAE,MAAM,EAAE,cAAc,CAAC,CAAC,CAAC,MAAM,EAAE,YAAY,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,KAAK,EAAE,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC;QAC3G,CAAC,CAAC,CAAC;QACH,MAAM,GAAG,GAAG;YACV,GAAG,OAAO,CAAC,GAAG,CAAC,UAAU,CAAC;YAC1B,mBAAmB;YACnB,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,EAAE,CAAC,WAAW,EAAE,IAAI,UAAU,CAAC,CAAC,CAAC,MAAM,CAAC,QAAQ,UAAU,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,CAAC;SAChG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACb,MAAM,WAAW,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,aAAa,OAAO,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC5F,qFAAqF;QACrF,MAAM,SAAS,GAAG,IAAI,GAAG,CAAC,CAAC,OAAO,EAAE,GAAG,OAAO,EAAE,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;QAC/E,MAAM,QAAQ,GAAG,KAAK,CAAC,OAAO,IAAI,OAAO,CAAC;QAC1C,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,QAAQ,CAAC;YAAE,MAAM,IAAI,KAAK,CAAC,YAAY,QAAQ,qBAAqB,CAAC,GAAG,SAAS,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QACpH,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,KAAK,IAAI,EAAE,EAAE,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC;QAC5D,MAAM,GAAG,GAAG,UAAU,GAAG,IAAI,IAAI,GAAG,WAAW,aAAa,UAAU,CAAC,QAAQ,CAAC,IAAI,QAAQ,UAAU,KAAK,EAAE,CAAC;QAC9G,MAAM,MAAM,GAAG,EAAE,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,MAAM,CAA8B,CAAC;QAC3E,MAAM,WAAW,GAAG,OAAO,CAAC,MAAM;YAChC,CAAC,CAAE,EAAE,CAAC,OAAO,CAAC,oCAAoC,IAAI,GAAG,WAAW,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,MAAM,CAAmB,CAAC,CAAC;YAC3G,CAAC,CAAC,CAAC,CAAC;QAEN,iDAAiD;QACjD,MAAM,GAAG,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE;YACzB,MAAM,GAAG,GAA4B,EAAE,CAAC;YACxC,KAAK,MAAM,CAAC,IAAI,OAAO;gBAAE,GAAG,CAAC,CAAC,CAAC,GAAG,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACrD,MAAM,KAAK,GAAG,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;YACnC,GAAG,CAAC,KAAK,GAAG,KAAK,CAAC;YAClB,GAAG,CAAC,GAAG,GAAG,SAAS,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,KAAK,GAAG,SAAS,CAAC,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;YAC1E,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;gBACxB,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC;gBACrB,GAAG,CAAC,CAAC,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,EAAE,KAAK,KAAK,IAAI,OAAO,CAAC,KAAK,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;YACzF,CAAC;YACD,OAAO,GAAG,CAAC;QACb,CAAC,CAAC,CAAC;QAEH,MAAM,UAAU,GAAG,CAAC,GAAG,OAAO,EAAE,OAAO,EAAE,GAAG,EAAE,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC;QAC5E,MAAM,MAAM,GAAG,KAAK,UAAU,CAAC,IAAI,CAAC,KAAK,CAAC,QAAQ,UAAU,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;QAC3F,MAAM,IAAI,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CACvB,KAAK,CAAC,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,EAAE,GAAG,CAAC,CAAC,GAAG,GAAG,EAAE,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAChI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACb,MAAM,QAAQ,GAAG,SAAS,GAAG,CAAC;YAC5B,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAI,CAAC,CAAC,KAAgB,EAAE,CAAC,CAAC,GAAG,SAAS,CAAC,GAAG,IAAI,CAAC,GAAG,EAAE;YACxF,CAAC,CAAC,CAAC,CAAC;QACN,MAAM,SAAS,GAAG,OAAO,CAAC,MAAM;YAC9B,CAAC,CAAC,UAAU,SAAS,CAAC,cAAc,CAAC,OAAO,CAAC,gBAAgB,WAAW,CAAC,cAAc,CAAC,OAAO,CAAC,qBAAqB,GAAG,CAAC,MAAM,GAAG,WAAW,GAAG,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,cAAc,QAAQ,qDAAqD,CAAC,CAAC,CAAC,EAAE,GAAG;YAChP,CAAC,CAAC,UAAU,SAAS,CAAC,cAAc,CAAC,OAAO,CAAC,wCAAwC,CAAC;QACxF,MAAM,QAAQ,GAAG,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,GAAG,MAAM,KAAK,IAAI,OAAO,SAAS,EAAE,CAAC,CAAC,CAAC,kBAAkB,SAAS,EAAE,CAAC;QACnG,OAAO;YACL,QAAQ;YACR,UAAU,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE,OAAO,EAAE,OAAO,EAAE,OAAO,CAAC,GAAG,CAAC,CAAC,EAAE,EAAE,EAAE,MAAM,EAAE,EAAE,EAAE,CAAC,CAAC,EAAE,EAAE,EAAE,MAAM,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,GAAG,EAAE,WAAW,EAAE,cAAc,EAAE,GAAG,CAAC,MAAM,EAAE,SAAS,EAAE;SAClK,CAAC;IACJ,CAAC;IAED,6FAA6F;IAC7F,MAAM,IAAI,GAAG,CAAC,KAAK,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,mBAAmB,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,UAAU,CAAC,CAAC,EAAE,SAAS,CAAC,CAAC,CAAC;IACrH,MAAM,QAAQ,GAAG,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,UAAU,CAAC,KAAK,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;IAC7E,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,KAAK,IAAI,EAAE,EAAE,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC;IAC5D,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,MAAM,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC;IAC9C,MAAM,GAAG,GAAG,UAAU,IAAI,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,IAAI,GAAG,QAAQ,CAAC,CAAC,CAAC,aAAa,UAAU,CAAC,QAAQ,CAAC,IAAI,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,UAAU,KAAK,WAAW,MAAM,EAAE,CAAC;IAClK,MAAM,GAAG,GAAG,EAAE,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,MAAM,CAA8B,CAAC;IACxE,IAAI,SAAS,GAAG,KAAK,CAAC;IACtB,MAAM,IAAI,GAAG,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE;QACvB,MAAM,CAAC,GAA4B,EAAE,CAAC;QACtC,KAAK,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC;YACrB,MAAM,CAAC,GAAG,YAAY,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YAC7B,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;gBAAE,SAAS,GAAG,IAAI,CAAC;YACjC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QACX,CAAC;QACD,OAAO,CAAC,CAAC;IACX,CAAC,CAAC,CAAC;IACH,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC3C,MAAM,IAAI,GAAG,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC;IAClC,MAAM,UAAU,GAAG,IAAI,GAAG,SAAS,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC;IAClD,MAAM,MAAM,GAAG,WAAW,KAAK,IAAI,IAAI,OAAO,SAAS,CAAC,cAAc,CAAC,OAAO,CAAC,GAAG,UAAU,IAAI,IAAI,CAAC,CAAC,CAAC,iBAAiB,UAAU,EAAE,CAAC,CAAC,CAAC,QAAQ,GAAG;QAChJ,CAAC,SAAS,CAAC,CAAC,CAAC,eAAe,QAAQ,qFAAqF,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;IAClI,MAAM,MAAM,GAAG,KAAK,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,QAAQ,IAAI,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;IAC/E,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,KAAK,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACxF,MAAM,QAAQ,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,GAAG,MAAM,KAAK,IAAI,OAAO,MAAM,EAAE,CAAC,CAAC,CAAC,kBAAkB,SAAS,aAAa,KAAK,sBAAsB,CAAC;IACvI,OAAO;QACL,QAAQ;QACR,UAAU,EAAE,EAAE,KAAK,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,IAAI,EAAE,YAAY,EAAE,IAAI,CAAC,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,UAAU,EAAE,cAAc,EAAE,SAAS,EAAE;KACtI,CAAC;AACJ,CAAC"}

package/dist/core/reranker.d.ts

@@ -0,0 +1,8 @@
+/** Relevance logits for one query against many passages (batched). Higher = more relevant. */
+export declare function scorePassages(query: string, passages: string[]): Promise<number[]>;
+/** The single best passage's score + its index — the "max-passage" relevance of a doc to a query. */
+export declare function maxPassageScore(query: string, passages: string[]): Promise<{
+    max: number;
+    index: number;
+}>;
+//# sourceMappingURL=reranker.d.ts.map

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

@@ -0,0 +1 @@
+{"version":3,"file":"reranker.d.ts","sourceRoot":"","sources":["../../src/core/reranker.ts"],"names":[],"mappings":"AA6CA,8FAA8F;AAC9F,wBAAsB,aAAa,CAAC,KAAK,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,CAYxF;AAED,qGAAqG;AACrG,wBAAsB,eAAe,CAAC,KAAK,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC;IAAE,GAAG,EAAE,MAAM,CAAC;IAAC,KAAK,EAAE,MAAM,CAAA;CAAE,CAAC,CAKhH"}

package/dist/core/reranker.js

@@ -0,0 +1,68 @@
+import { env, AutoTokenizer, AutoModelForSequenceClassification } from '@huggingface/transformers';
+import { homedir } from 'node:os';
+import path from 'node:path';
+// Cross-encoder reranker (ms-marco-MiniLM-L-6-v2) run locally via transformers.js / ONNX — pure Node,
+// no Python. It SCORES (query, passage) relevance; it does not generate text, so it cannot hallucinate.
+// Model is downloaded once from the HF hub and cached on disk (persists across npm reinstalls).
+env.cacheDir = path.join(homedir(), '.cache', 'seo-audit-console-models');
+env.allowLocalModels = false;
+const MODEL = 'Xenova/ms-marco-MiniLM-L-6-v2';
+let loading = null;
+// Prefer the GPU when present. onnxruntime-node bundles the DirectML provider (any DX12 GPU on
+// Windows — incl. NVIDIA), not the CUDA EP, so 'dml' is the GPU path; fall back to CPU if it won't
+// init. (mirrors ai-detect's `cuda if available else cpu`, adapted to our ONNX runtime). Override
+// with SAC_RERANK_DEVICE=cpu|dml. fp32 — DirectML doesn't run the q8-quantised graph.
+const DEVICE_ORDER = process.env.SAC_RERANK_DEVICE
+    ? [process.env.SAC_RERANK_DEVICE]
+    : ['dml', 'cpu'];
+/** Lazy singleton — first call downloads (~25MB) + warms the model on the best device; later calls reuse it. */
+function load() {
+    if (!loading) {
+        loading = (async () => {
+            const tokenizer = await AutoTokenizer.from_pretrained(MODEL);
+            for (let i = 0; i < DEVICE_ORDER.length; i++) {
+                const device = DEVICE_ORDER[i];
+                try {
+                    const model = await AutoModelForSequenceClassification.from_pretrained(MODEL, { device: device, dtype: 'fp32' });
+                    console.error(`[reranker] ms-marco-MiniLM loaded on ${device}`);
+                    return { tokenizer, model, device };
+                }
+                catch (e) {
+                    const last = i === DEVICE_ORDER.length - 1;
+                    console.error(`[reranker] device '${device}' unavailable${last ? '' : ', falling back'}: ${e instanceof Error ? e.message.slice(0, 100) : e}`);
+                    if (last)
+                        throw e;
+                }
+            }
+            throw new Error('reranker: no usable device');
+        })();
+    }
+    return loading;
+}
+/** Relevance logits for one query against many passages (batched). Higher = more relevant. */
+export async function scorePassages(query, passages) {
+    if (!passages.length || !query.trim())
+        return [];
+    const { tokenizer, model } = await load();
+    const out = [];
+    const BATCH = 16; // bound peak memory
+    for (let i = 0; i < passages.length; i += BATCH) {
+        const batch = passages.slice(i, i + BATCH);
+        const inputs = tokenizer(Array(batch.length).fill(query), { text_pair: batch, padding: true, truncation: true });
+        const { logits } = await model(inputs);
+        for (const row of logits.tolist())
+            out.push(row[0]);
+    }
+    return out;
+}
+/** The single best passage's score + its index — the "max-passage" relevance of a doc to a query. */
+export async function maxPassageScore(query, passages) {
+    const scores = await scorePassages(query, passages);
+    let max = -Infinity, index = -1;
+    scores.forEach((v, i) => { if (v > max) {
+        max = v;
+        index = i;
+    } });
+    return { max, index };
+}
+//# sourceMappingURL=reranker.js.map

package/dist/core/reranker.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"reranker.js","sourceRoot":"","sources":["../../src/core/reranker.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,GAAG,EAAE,aAAa,EAAE,kCAAkC,EAAE,MAAM,2BAA2B,CAAC;AACnG,OAAO,EAAE,OAAO,EAAE,MAAM,SAAS,CAAC;AAClC,OAAO,IAAI,MAAM,WAAW,CAAC;AAE7B,sGAAsG;AACtG,wGAAwG;AACxG,gGAAgG;AAChG,GAAG,CAAC,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,EAAE,EAAE,QAAQ,EAAE,0BAA0B,CAAC,CAAC;AAC1E,GAAG,CAAC,gBAAgB,GAAG,KAAK,CAAC;AAE7B,MAAM,KAAK,GAAG,+BAA+B,CAAC;AAE9C,IAAI,OAAO,GAA2B,IAAI,CAAC;AAE3C,+FAA+F;AAC/F,mGAAmG;AACnG,kGAAkG;AAClG,sFAAsF;AACtF,MAAM,YAAY,GAAa,OAAO,CAAC,GAAG,CAAC,iBAAiB;IAC1D,CAAC,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,iBAAiB,CAAC;IACjC,CAAC,CAAC,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC;AAEnB,gHAAgH;AAChH,SAAS,IAAI;IACX,IAAI,CAAC,OAAO,EAAE,CAAC;QACb,OAAO,GAAG,CAAC,KAAK,IAAI,EAAE;YACpB,MAAM,SAAS,GAAG,MAAM,aAAa,CAAC,eAAe,CAAC,KAAK,CAAC,CAAC;YAC7D,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,YAAY,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;gBAC7C,MAAM,MAAM,GAAG,YAAY,CAAC,CAAC,CAAC,CAAC;gBAC/B,IAAI,CAAC;oBACH,MAAM,KAAK,GAAG,MAAM,kCAAkC,CAAC,eAAe,CAAC,KAAK,EAAE,EAAE,MAAM,EAAE,MAAyC,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC,CAAC;oBACpJ,OAAO,CAAC,KAAK,CAAC,wCAAwC,MAAM,EAAE,CAAC,CAAC;oBAChE,OAAO,EAAE,SAAS,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC;gBACtC,CAAC;gBAAC,OAAO,CAAC,EAAE,CAAC;oBACX,MAAM,IAAI,GAAG,CAAC,KAAK,YAAY,CAAC,MAAM,GAAG,CAAC,CAAC;oBAC3C,OAAO,CAAC,KAAK,CAAC,sBAAsB,MAAM,gBAAgB,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,gBAAgB,KAAK,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;oBAC/I,IAAI,IAAI;wBAAE,MAAM,CAAC,CAAC;gBACpB,CAAC;YACH,CAAC;YACD,MAAM,IAAI,KAAK,CAAC,4BAA4B,CAAC,CAAC;QAChD,CAAC,CAAC,EAAE,CAAC;IACP,CAAC;IACD,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,8FAA8F;AAC9F,MAAM,CAAC,KAAK,UAAU,aAAa,CAAC,KAAa,EAAE,QAAkB;IACnE,IAAI,CAAC,QAAQ,CAAC,MAAM,IAAI,CAAC,KAAK,CAAC,IAAI,EAAE;QAAE,OAAO,EAAE,CAAC;IACjD,MAAM,EAAE,SAAS,EAAE,KAAK,EAAE,GAAG,MAAM,IAAI,EAAE,CAAC;IAC1C,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,MAAM,KAAK,GAAG,EAAE,CAAC,CAAC,oBAAoB;IACtC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,MAAM,EAAE,CAAC,IAAI,KAAK,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,QAAQ,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,CAAC;QAC3C,MAAM,MAAM,GAAG,SAAS,CAAC,KAAK,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,EAAE,SAAS,EAAE,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,UAAU,EAAE,IAAI,EAAE,CAAC,CAAC;QACjH,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,KAAK,CAAC,MAAM,CAAC,CAAC;QACvC,KAAK,MAAM,GAAG,IAAI,MAAM,CAAC,MAAM,EAAgB;YAAE,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IACpE,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,qGAAqG;AACrG,MAAM,CAAC,KAAK,UAAU,eAAe,CAAC,KAAa,EAAE,QAAkB;IACrE,MAAM,MAAM,GAAG,MAAM,aAAa,CAAC,KAAK,EAAE,QAAQ,CAAC,CAAC;IACpD,IAAI,GAAG,GAAG,CAAC,QAAQ,EAAE,KAAK,GAAG,CAAC,CAAC,CAAC;IAChC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,GAAG,IAAI,CAAC,GAAG,GAAG,EAAE,CAAC;QAAC,GAAG,GAAG,CAAC,CAAC;QAAC,KAAK,GAAG,CAAC,CAAC;IAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACnE,OAAO,EAAE,GAAG,EAAE,KAAK,EAAE,CAAC;AACxB,CAAC"}

package/dist/core/robots.d.ts

@@ -0,0 +1,11 @@
+/**
+ * robots.txt politeness gate. Honours both Disallow AND Allow with longest-prefix-match
+ * (most specific path wins; Allow wins ties) — so a broad `Disallow: /` paired with
+ * `Allow: /public/` doesn't wrongly empty the crawl. NB: prefix match only — `*`/`$`
+ * wildcard patterns (e.g. `Disallow: /*?`) are not yet interpreted (owned-site crawl).
+ */
+export interface RobotsRules {
+    isAllowed(pathname: string): boolean;
+}
+export declare function fetchRobots(origin: string, userAgent: string, uaToken?: string): Promise<RobotsRules>;
+//# sourceMappingURL=robots.d.ts.map

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

@@ -0,0 +1 @@
+{"version":3,"file":"robots.d.ts","sourceRoot":"","sources":["../../src/core/robots.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AACH,MAAM,WAAW,WAAW;IAC1B,SAAS,CAAC,QAAQ,EAAE,MAAM,GAAG,OAAO,CAAC;CACtC;AA0DD,wBAAsB,WAAW,CAAC,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,OAAO,SAAsB,GAAG,OAAO,CAAC,WAAW,CAAC,CAWxH"}

package/dist/core/robots.js

@@ -0,0 +1,76 @@
+const ALLOW_ALL = { isAllowed: () => true };
+function parseRules(txt, uaToken) {
+    const groups = [];
+    let current = null;
+    let lastWasAgent = false;
+    for (const raw of txt.split(/\r?\n/)) {
+        const line = raw.replace(/#.*$/, '').trim();
+        if (!line)
+            continue;
+        const idx = line.indexOf(':');
+        if (idx === -1)
+            continue;
+        const field = line.slice(0, idx).trim().toLowerCase();
+        const value = line.slice(idx + 1).trim();
+        if (field === 'user-agent') {
+            if (!current || !lastWasAgent) {
+                current = { agents: [], rules: [] };
+                groups.push(current);
+            }
+            current.agents.push(value.toLowerCase());
+            lastWasAgent = true;
+        }
+        else if ((field === 'disallow' || field === 'allow') && current) {
+            // An empty Disallow value means "allow everything" — carries no restriction.
+            if (value)
+                current.rules.push({ path: value, allow: field === 'allow' });
+            lastWasAgent = false;
+        }
+        else {
+            lastWasAgent = false;
+        }
+    }
+    // RFC 9309 §2.2.1: only the MOST SPECIFIC matching group applies — a bot-specific group
+    // replaces the `*` group entirely (it does not merge with it). Otherwise a site that
+    // whitelists this bot (`User-agent: seo-audit-console / Allow: /`) alongside a broad
+    // `User-agent: * / Disallow: /` would still be blocked. Groups at equal specificity merge.
+    const ua = uaToken.toLowerCase();
+    const agentScore = (a) => (a === '*' ? 0 : ua.includes(a) ? a.length : -1);
+    const groupScore = (g) => Math.max(-1, ...g.agents.map(agentScore));
+    const best = Math.max(-1, ...groups.map(groupScore));
+    if (best < 0)
+        return [];
+    return groups.filter(g => groupScore(g) === best).flatMap(g => g.rules);
+}
+function makeRules(rules) {
+    if (!rules.length)
+        return ALLOW_ALL;
+    return {
+        isAllowed(p) {
+            let best = null;
+            for (const r of rules) {
+                if (!p.startsWith(r.path))
+                    continue;
+                // longest match wins; Allow beats Disallow at equal length
+                if (!best || r.path.length > best.path.length || (r.path.length === best.path.length && r.allow))
+                    best = r;
+            }
+            return best ? best.allow : true;
+        },
+    };
+}
+export async function fetchRobots(origin, userAgent, uaToken = 'seo-audit-console') {
+    try {
+        const res = await fetch(new URL('/robots.txt', origin), {
+            headers: { 'user-agent': userAgent, 'cache-control': 'no-cache' },
+            signal: AbortSignal.timeout(10000),
+        });
+        if (!res.ok)
+            return ALLOW_ALL;
+        return makeRules(parseRules(await res.text(), uaToken));
+    }
+    catch {
+        return ALLOW_ALL; // robots unavailable → don't block (owned-site crawl)
+    }
+}
+//# sourceMappingURL=robots.js.map

package/dist/core/robots.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"robots.js","sourceRoot":"","sources":["../../src/core/robots.ts"],"names":[],"mappings":"AAWA,MAAM,SAAS,GAAgB,EAAE,SAAS,EAAE,GAAG,EAAE,CAAC,IAAI,EAAE,CAAC;AAEzD,SAAS,UAAU,CAAC,GAAW,EAAE,OAAe;IAC9C,MAAM,MAAM,GAA0C,EAAE,CAAC;IACzD,IAAI,OAAO,GAA+C,IAAI,CAAC;IAC/D,IAAI,YAAY,GAAG,KAAK,CAAC;IAEzB,KAAK,MAAM,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,OAAO,CAAC,EAAE,CAAC;QACrC,MAAM,IAAI,GAAG,GAAG,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;QAC5C,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,CAAC;QAC9B,IAAI,GAAG,KAAK,CAAC,CAAC;YAAE,SAAS;QACzB,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;QACtD,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAEzC,IAAI,KAAK,KAAK,YAAY,EAAE,CAAC;YAC3B,IAAI,CAAC,OAAO,IAAI,CAAC,YAAY,EAAE,CAAC;gBAAC,OAAO,GAAG,EAAE,MAAM,EAAE,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE,CAAC;gBAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YAAC,CAAC;YAC7F,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,WAAW,EAAE,CAAC,CAAC;YACzC,YAAY,GAAG,IAAI,CAAC;QACtB,CAAC;aAAM,IAAI,CAAC,KAAK,KAAK,UAAU,IAAI,KAAK,KAAK,OAAO,CAAC,IAAI,OAAO,EAAE,CAAC;YAClE,6EAA6E;YAC7E,IAAI,KAAK;gBAAE,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,KAAK,OAAO,EAAE,CAAC,CAAC;YACzE,YAAY,GAAG,KAAK,CAAC;QACvB,CAAC;aAAM,CAAC;YACN,YAAY,GAAG,KAAK,CAAC;QACvB,CAAC;IACH,CAAC;IAED,wFAAwF;IACxF,qFAAqF;IACrF,qFAAqF;IACrF,2FAA2F;IAC3F,MAAM,EAAE,GAAG,OAAO,CAAC,WAAW,EAAE,CAAC;IACjC,MAAM,UAAU,GAAG,CAAC,CAAS,EAAU,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC3F,MAAM,UAAU,GAAG,CAAC,CAAuB,EAAU,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC;IAClG,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,GAAG,MAAM,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC;IACrD,IAAI,IAAI,GAAG,CAAC;QAAE,OAAO,EAAE,CAAC;IACxB,OAAO,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,UAAU,CAAC,CAAC,CAAC,KAAK,IAAI,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC;AAC1E,CAAC;AAED,SAAS,SAAS,CAAC,KAAa;IAC9B,IAAI,CAAC,KAAK,CAAC,MAAM;QAAE,OAAO,SAAS,CAAC;IACpC,OAAO;QACL,SAAS,CAAC,CAAS;YACjB,IAAI,IAAI,GAAgB,IAAI,CAAC;YAC7B,KAAK,MAAM,CAAC,IAAI,KAAK,EAAE,CAAC;gBACtB,IAAI,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,CAAC;oBAAE,SAAS;gBACpC,2DAA2D;gBAC3D,IAAI,CAAC,IAAI,IAAI,CAAC,CAAC,IAAI,CAAC,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,KAAK,IAAI,CAAC,IAAI,CAAC,MAAM,IAAI,CAAC,CAAC,KAAK,CAAC;oBAAE,IAAI,GAAG,CAAC,CAAC;YAC7G,CAAC;YACD,OAAO,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC;QAClC,CAAC;KACF,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,WAAW,CAAC,MAAc,EAAE,SAAiB,EAAE,OAAO,GAAG,mBAAmB;IAChG,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,IAAI,GAAG,CAAC,aAAa,EAAE,MAAM,CAAC,EAAE;YACtD,OAAO,EAAE,EAAE,YAAY,EAAE,SAAS,EAAE,eAAe,EAAE,UAAU,EAAE;YACjE,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,KAAK,CAAC;SACnC,CAAC,CAAC;QACH,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,OAAO,SAAS,CAAC;QAC9B,OAAO,SAAS,CAAC,UAAU,CAAC,MAAM,GAAG,CAAC,IAAI,EAAE,EAAE,OAAO,CAAC,CAAC,CAAC;IAC1D,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,SAAS,CAAC,CAAC,sDAAsD;IAC1E,CAAC;AACH,CAAC"}

package/dist/core/sitemap.d.ts

@@ -0,0 +1,15 @@
+import { type UrlKeyOptions } from './url-key.js';
+export interface SitemapResult {
+    urls: {
+        urlKey: string;
+        url: string;
+        lastmod: string | null;
+    }[];
+    sitemapsFetched: number;
+    sources: string[];
+}
+export declare function fetchSitemapUrls(origin: string, ua: string, keyOpts: UrlKeyOptions, opts?: {
+    maxSitemaps?: number;
+    maxUrls?: number;
+}): Promise<SitemapResult>;
+//# sourceMappingURL=sitemap.d.ts.map

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

@@ -0,0 +1 @@
+{"version":3,"file":"sitemap.d.ts","sourceRoot":"","sources":["../../src/core/sitemap.ts"],"names":[],"mappings":"AACA,OAAO,EAAU,KAAK,aAAa,EAAE,MAAM,cAAc,CAAC;AAiC1D,MAAM,WAAW,aAAa;IAAG,IAAI,EAAE;QAAE,MAAM,EAAE,MAAM,CAAC;QAAC,GAAG,EAAE,MAAM,CAAC;QAAC,OAAO,EAAE,MAAM,GAAG,IAAI,CAAA;KAAE,EAAE,CAAC;IAAC,eAAe,EAAE,MAAM,CAAC;IAAC,OAAO,EAAE,MAAM,EAAE,CAAA;CAAE;AAE9I,wBAAsB,gBAAgB,CACpC,MAAM,EAAE,MAAM,EAAE,EAAE,EAAE,MAAM,EAAE,OAAO,EAAE,aAAa,EAAE,IAAI,GAAE;IAAE,WAAW,CAAC,EAAE,MAAM,CAAC;IAAC,OAAO,CAAC,EAAE,MAAM,CAAA;CAAO,GACxG,OAAO,CAAC,aAAa,CAAC,CAgDxB"}

package/dist/core/sitemap.js

@@ -0,0 +1,100 @@
+import { gunzipSync } from 'node:zlib';
+import { urlKey } from './url-key.js';
+/**
+ * Sitemap discovery + parse for reconciliation against the crawl. Finds sitemaps via robots.txt
+ * `Sitemap:` directives (falling back to /sitemap.xml), follows <sitemapindex> children, handles
+ * gzipped sitemaps, and returns the listed URLs normalised to url_key. Bounded (count + timeout)
+ * so a giant sitemap set can't stall a crawl.
+ */
+const LOC = /<loc>\s*([^<\s]+?)\s*<\/loc>/gi;
+// XML sitemaps legally escape URL characters (?a=1&b=2 is stored as ?a=1&amp;b=2). Decode
+// numeric + the five predefined entities so stored URLs match crawl/GSC url_keys and the
+// crawler seeds fetch the real URL, not the literal `&amp;` form. `&amp;` is decoded last
+// so `&amp;lt;` correctly yields `&lt;` rather than `<`.
+const decodeXmlEntities = (s) => s
+    .replace(/&#x([0-9a-f]+);/gi, (_, h) => String.fromCodePoint(parseInt(h, 16)))
+    .replace(/&#(\d+);/g, (_, d) => String.fromCodePoint(Number(d)))
+    .replace(/&lt;/g, '<').replace(/&gt;/g, '>').replace(/&quot;/g, '"').replace(/&apos;/g, "'")
+    .replace(/&amp;/g, '&');
+async function fetchText(url, ua) {
+    try {
+        const res = await fetch(url, { headers: { 'user-agent': ua }, signal: AbortSignal.timeout(15000) });
+        if (!res.ok)
+            return null;
+        const buf = Buffer.from(await res.arrayBuffer());
+        const ct = (res.headers.get('content-type') ?? '').toLowerCase();
+        if (url.toLowerCase().endsWith('.gz') || ct.includes('gzip')) {
+            try {
+                return gunzipSync(buf).toString('utf8');
+            }
+            catch {
+                return buf.toString('utf8');
+            }
+        }
+        return buf.toString('utf8');
+    }
+    catch {
+        return null;
+    }
+}
+export async function fetchSitemapUrls(origin, ua, keyOpts, opts = {}) {
+    const maxSitemaps = opts.maxSitemaps ?? 60;
+    const maxUrls = opts.maxUrls ?? 100000;
+    // Discover sitemap locations from robots.txt; fall back to /sitemap.xml.
+    const sources = [];
+    const robots = await fetchText(new URL('/robots.txt', origin).toString(), ua);
+    if (robots)
+        for (const m of robots.matchAll(/^\s*sitemap:\s*(\S+)/gim))
+            sources.push(m[1].trim());
+    if (!sources.length)
+        sources.push(new URL('/sitemap.xml', origin).toString());
+    const out = new Map(); // urlKey → entry (dedup)
+    const queue = [...new Set(sources)];
+    let fetched = 0;
+    const URL_BLOCK = /<url>([\s\S]*?)<\/url>/gi;
+    const LOC_ONE = /<loc>\s*([^<\s]+?)\s*<\/loc>/i;
+    const LASTMOD_ONE = /<lastmod>\s*([^<\s]+?)\s*<\/lastmod>/i;
+    const add = (loc, lastmod) => {
+        try {
+            const k = urlKey(loc, keyOpts);
+            if (!out.has(k))
+                out.set(k, { url: loc, lastmod });
+        }
+        catch { /* skip bad loc */ }
+    };
+    while (queue.length && fetched < maxSitemaps && out.size < maxUrls) {
+        const sm = queue.shift();
+        const txt = await fetchText(sm, ua);
+        fetched++;
+        if (!txt)
+            continue;
+        if (/<sitemapindex[\s>]/i.test(txt)) {
+            for (const m of txt.matchAll(LOC)) {
+                if (queue.length + fetched < maxSitemaps)
+                    queue.push(decodeXmlEntities(m[1])); // child sitemap
+            }
+            continue;
+        }
+        // Parse per-<url> block so each <lastmod> claim stays paired with its <loc>.
+        let sawBlock = false;
+        for (const b of txt.matchAll(URL_BLOCK)) {
+            sawBlock = true;
+            const loc = b[1].match(LOC_ONE)?.[1];
+            if (!loc)
+                continue;
+            add(decodeXmlEntities(loc), b[1].match(LASTMOD_ONE)?.[1] ?? null);
+            if (out.size >= maxUrls)
+                break;
+        }
+        // Defensive fallback for malformed sitemaps that list bare <loc>s without <url> wrappers.
+        if (!sawBlock) {
+            for (const m of txt.matchAll(LOC)) {
+                add(decodeXmlEntities(m[1]), null);
+                if (out.size >= maxUrls)
+                    break;
+            }
+        }
+    }
+    return { urls: [...out].map(([k, v]) => ({ urlKey: k, url: v.url, lastmod: v.lastmod })), sitemapsFetched: fetched, sources };
+}
+//# sourceMappingURL=sitemap.js.map

package/dist/core/sitemap.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"sitemap.js","sourceRoot":"","sources":["../../src/core/sitemap.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,WAAW,CAAC;AACvC,OAAO,EAAE,MAAM,EAAsB,MAAM,cAAc,CAAC;AAE1D;;;;;GAKG;AACH,MAAM,GAAG,GAAG,gCAAgC,CAAC;AAE7C,0FAA0F;AAC1F,yFAAyF;AACzF,0FAA0F;AAC1F,yDAAyD;AACzD,MAAM,iBAAiB,GAAG,CAAC,CAAS,EAAU,EAAE,CAAC,CAAC;KAC/C,OAAO,CAAC,mBAAmB,EAAE,CAAC,CAAC,EAAE,CAAS,EAAE,EAAE,CAAC,MAAM,CAAC,aAAa,CAAC,QAAQ,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC;KACrF,OAAO,CAAC,WAAW,EAAE,CAAC,CAAC,EAAE,CAAS,EAAE,EAAE,CAAC,MAAM,CAAC,aAAa,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC;KACvE,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;KAC3F,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;AAE1B,KAAK,UAAU,SAAS,CAAC,GAAW,EAAE,EAAU;IAC9C,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,GAAG,EAAE,EAAE,OAAO,EAAE,EAAE,YAAY,EAAE,EAAE,EAAE,EAAE,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;QACpG,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,OAAO,IAAI,CAAC;QACzB,MAAM,GAAG,GAAG,MAAM,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,WAAW,EAAE,CAAC,CAAC;QACjD,MAAM,EAAE,GAAG,CAAC,GAAG,CAAC,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC,CAAC,WAAW,EAAE,CAAC;QACjE,IAAI,GAAG,CAAC,WAAW,EAAE,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,EAAE,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;YAC7D,IAAI,CAAC;gBAAC,OAAO,UAAU,CAAC,GAAG,CAAC,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;YAAC,CAAC;YAAC,MAAM,CAAC;gBAAC,OAAO,GAAG,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;YAAC,CAAC;QACzF,CAAC;QACD,OAAO,GAAG,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC;IAC9B,CAAC;IAAC,MAAM,CAAC;QAAC,OAAO,IAAI,CAAC;IAAC,CAAC;AAC1B,CAAC;AAID,MAAM,CAAC,KAAK,UAAU,gBAAgB,CACpC,MAAc,EAAE,EAAU,EAAE,OAAsB,EAAE,OAAmD,EAAE;IAEzG,MAAM,WAAW,GAAG,IAAI,CAAC,WAAW,IAAI,EAAE,CAAC;IAC3C,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,IAAI,MAAM,CAAC;IAEvC,yEAAyE;IACzE,MAAM,OAAO,GAAa,EAAE,CAAC;IAC7B,MAAM,MAAM,GAAG,MAAM,SAAS,CAAC,IAAI,GAAG,CAAC,aAAa,EAAE,MAAM,CAAC,CAAC,QAAQ,EAAE,EAAE,EAAE,CAAC,CAAC;IAC9E,IAAI,MAAM;QAAE,KAAK,MAAM,CAAC,IAAI,MAAM,CAAC,QAAQ,CAAC,yBAAyB,CAAC;YAAE,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;IAClG,IAAI,CAAC,OAAO,CAAC,MAAM;QAAE,OAAO,CAAC,IAAI,CAAC,IAAI,GAAG,CAAC,cAAc,EAAE,MAAM,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC;IAE9E,MAAM,GAAG,GAAG,IAAI,GAAG,EAAmD,CAAC,CAAC,yBAAyB;IACjG,MAAM,KAAK,GAAG,CAAC,GAAG,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC;IACpC,IAAI,OAAO,GAAG,CAAC,CAAC;IAChB,MAAM,SAAS,GAAG,0BAA0B,CAAC;IAC7C,MAAM,OAAO,GAAG,+BAA+B,CAAC;IAChD,MAAM,WAAW,GAAG,uCAAuC,CAAC;IAC5D,MAAM,GAAG,GAAG,CAAC,GAAW,EAAE,OAAsB,EAAQ,EAAE;QACxD,IAAI,CAAC;YAAC,MAAM,CAAC,GAAG,MAAM,CAAC,GAAG,EAAE,OAAO,CAAC,CAAC;YAAC,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC;gBAAE,GAAG,CAAC,GAAG,CAAC,CAAC,EAAE,EAAE,GAAG,EAAE,GAAG,EAAE,OAAO,EAAE,CAAC,CAAC;QAAC,CAAC;QAAC,MAAM,CAAC,CAAC,kBAAkB,CAAC,CAAC;IAC1H,CAAC,CAAC;IACF,OAAO,KAAK,CAAC,MAAM,IAAI,OAAO,GAAG,WAAW,IAAI,GAAG,CAAC,IAAI,GAAG,OAAO,EAAE,CAAC;QACnE,MAAM,EAAE,GAAG,KAAK,CAAC,KAAK,EAAG,CAAC;QAC1B,MAAM,GAAG,GAAG,MAAM,SAAS,CAAC,EAAE,EAAE,EAAE,CAAC,CAAC;QACpC,OAAO,EAAE,CAAC;QACV,IAAI,CAAC,GAAG;YAAE,SAAS;QACnB,IAAI,qBAAqB,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YACpC,KAAK,MAAM,CAAC,IAAI,GAAG,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;gBAClC,IAAI,KAAK,CAAC,MAAM,GAAG,OAAO,GAAG,WAAW;oBAAE,KAAK,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,gBAAgB;YACjG,CAAC;YACD,SAAS;QACX,CAAC;QACD,6EAA6E;QAC7E,IAAI,QAAQ,GAAG,KAAK,CAAC;QACrB,KAAK,MAAM,CAAC,IAAI,GAAG,CAAC,QAAQ,CAAC,SAAS,CAAC,EAAE,CAAC;YACxC,QAAQ,GAAG,IAAI,CAAC;YAChB,MAAM,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;YACrC,IAAI,CAAC,GAAG;gBAAE,SAAS;YACnB,GAAG,CAAC,iBAAiB,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,WAAW,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;YAClE,IAAI,GAAG,CAAC,IAAI,IAAI,OAAO;gBAAE,MAAM;QACjC,CAAC;QACD,0FAA0F;QAC1F,IAAI,CAAC,QAAQ,EAAE,CAAC;YACd,KAAK,MAAM,CAAC,IAAI,GAAG,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;gBAClC,GAAG,CAAC,iBAAiB,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;gBACnC,IAAI,GAAG,CAAC,IAAI,IAAI,OAAO;oBAAE,MAAM;YACjC,CAAC;QACH,CAAC;IACH,CAAC;IACD,OAAO,EAAE,IAAI,EAAE,CAAC,GAAG,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,GAAG,EAAE,CAAC,CAAC,GAAG,EAAE,OAAO,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,EAAE,eAAe,EAAE,OAAO,EAAE,OAAO,EAAE,CAAC;AAChI,CAAC"}

package/dist/core/sql.d.ts

@@ -0,0 +1,6 @@
+/**
+ * Shared SQL fragments. There is exactly ONE definition of "is this page HTML" in SQL —
+ * keep it aligned with the runtime helper `isHtmlContentType` in Crawler.ts.
+ */
+export declare const HTML_CT = "(LOWER(content_type) LIKE 'text/html%' OR LOWER(content_type) LIKE 'application/xhtml+xml%')";
+//# sourceMappingURL=sql.d.ts.map

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

@@ -0,0 +1 @@
+{"version":3,"file":"sql.d.ts","sourceRoot":"","sources":["../../src/core/sql.ts"],"names":[],"mappings":"AAAA;;;GAGG;AACH,eAAO,MAAM,OAAO,iGAAiG,CAAC"}

package/dist/core/sql.js

@@ -0,0 +1,6 @@
+/**
+ * Shared SQL fragments. There is exactly ONE definition of "is this page HTML" in SQL —
+ * keep it aligned with the runtime helper `isHtmlContentType` in Crawler.ts.
+ */
+export const HTML_CT = `(LOWER(content_type) LIKE 'text/html%' OR LOWER(content_type) LIKE 'application/xhtml+xml%')`;
+//# sourceMappingURL=sql.js.map

package/dist/core/sql.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"sql.js","sourceRoot":"","sources":["../../src/core/sql.ts"],"names":[],"mappings":"AAAA;;;GAGG;AACH,MAAM,CAAC,MAAM,OAAO,GAAG,8FAA8F,CAAC"}

package/dist/core/types.d.ts

@@ -0,0 +1,22 @@
+export interface GscProperty {
+    siteUrl: string;
+    permissionLevel?: string;
+}
+export interface GscApiRow {
+    keys?: string[];
+    clicks: number;
+    impressions: number;
+    ctr: number;
+    position: number;
+}
+export interface FetchOptions {
+    startDate: string;
+    endDate: string;
+    dimensions?: string[];
+    rowLimit?: number;
+    dataState?: string;
+    searchType?: string;
+    fullResync?: boolean;
+}
+export type JobState = 'queued' | 'running' | 'completed' | 'failed' | 'cancelled';
+//# sourceMappingURL=types.d.ts.map