mcp-researchpowerpack 7.0.11 → 7.0.13

package/dist/src/utils/source-type.js

@@ -1,4 +1,5 @@
-const RULES = [
+// src/utils/source-type.ts
+var RULES = [
   // Reddit post permalinks (subreddit homepages are filtered out upstream).
   [/(?:^|\.)reddit\.com\//i, "reddit"],
   [/(?:^|\.)github\.com\//i, "github"],
@@ -35,7 +36,7 @@ function classifySourceByUrl(url) {
   }
   return "web";
 }
-const DOCUMENT_PATH_SUFFIXES = [
+var DOCUMENT_PATH_SUFFIXES = [
   ".pdf",
   ".doc",
   ".docx",
@@ -56,7 +57,7 @@ function isDocumentUrl(url) {
   }
   return false;
 }
-const BINARY_CONTENT_TYPE_PREFIXES = [
+var BINARY_CONTENT_TYPE_PREFIXES = [
   "application/pdf",
   "application/msword",
   "application/vnd.openxmlformats-officedocument.",

package/dist/src/utils/source-type.js.map

@@ -2,6 +2,6 @@
   "version": 3,
   "sources": ["../../../src/utils/source-type.ts"],
   "sourcesContent": ["/**\n * Hostname/path-heuristic source-type tagging. Works without the LLM\n * classifier so degraded-mode search responses still carry a\n * `source_type` field per result. When the LLM classifier IS available,\n * its tag wins (the classifier sees title + snippet as well, not just URL).\n *\n * See: mcp-revisions/output-shaping/06-source-type-tagging-without-llm.md.\n */\n\nexport type SourceType =\n  | 'reddit'\n  | 'github'\n  | 'docs'\n  | 'blog'\n  | 'paper'\n  | 'qa'\n  | 'cve'\n  | 'news'\n  | 'video'\n  | 'web';\n\nconst RULES: Array<[RegExp, SourceType]> = [\n  // Reddit post permalinks (subreddit homepages are filtered out upstream).\n  [/(?:^|\\.)reddit\\.com\\//i, 'reddit'],\n  [/(?:^|\\.)github\\.com\\//i, 'github'],\n  [/(?:^|\\.)gitlab\\.com\\//i, 'github'],\n  // CVE-prefixed paths are unambiguous regardless of host.\n  [/\\/CVE-\\d{4}-\\d+/i, 'cve'],\n  [/(?:^|\\.)nvd\\.nist\\.gov\\//i, 'cve'],\n  [/(?:^|\\.)stackoverflow\\.com\\//i, 'qa'],\n  [/(?:^|\\.)stackexchange\\.com\\//i, 'qa'],\n  [/(?:^|\\.)arxiv\\.org\\//i, 'paper'],\n  [/(?:^|\\.)medium\\.com\\//i, 'blog'],\n  [/(?:^|\\.)dev\\.to\\//i, 'blog'],\n  [/(?:^|\\.)substack\\.com\\//i, 'blog'],\n  // Docs subdomains and /docs/ paths.\n  [/^(?:[a-z0-9-]+\\.)*docs\\./i, 'docs'],\n  [/\\/docs\\//i, 'docs'],\n  [/(?:^|\\.)readthedocs\\.io\\//i, 'docs'],\n  // Video.\n  [/(?:^|\\.)youtube\\.com\\/watch/i, 'video'],\n  [/(?:^|\\.)youtu\\.be\\//i, 'video'],\n  // News / engineering blogs (last so it doesn't capture vendor docs).\n  [/(?:^|\\.)(?:news|blog|engineering)\\.[a-z0-9-]+\\.[a-z]{2,}\\//i, 'news'],\n];\n\nexport function classifySourceByUrl(url: string): SourceType {\n  let candidate: string;\n  try {\n    const u = new URL(url);\n    // Match against `host + pathname` so rules can use either or both.\n    candidate = `${u.hostname}${u.pathname}`;\n  } catch {\n    candidate = url;\n  }\n  for (const [re, type] of RULES) {\n    if (re.test(candidate)) return type;\n  }\n  return 'web';\n}\n\n// \u2500\u2500 Document-format detection (PDF / Office) \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\n// Scrape.do + Readability + Turndown assume HTML input. Binary document\n// formats need a markdown-extraction service (Jina Reader) instead. These two\n// helpers give the scrape pipeline both a pre-fetch gate (URL suffix) and a\n// post-fetch gate (response Content-Type header).\n\nconst DOCUMENT_PATH_SUFFIXES = [\n  '.pdf',\n  '.doc', '.docx',\n  '.ppt', '.pptx',\n  '.xls', '.xlsx',\n] as const;\n\n/**\n * Pre-fetch gate: does this URL's path end in a known binary-document suffix?\n * Case-insensitive. Trailing query strings / fragments are ignored \u2014 only the\n * pathname is inspected. Invalid URLs return false (handled upstream).\n */\nexport function isDocumentUrl(url: string): boolean {\n  let pathname: string;\n  try {\n    pathname = new URL(url).pathname.toLowerCase();\n  } catch {\n    return false;\n  }\n  for (const suffix of DOCUMENT_PATH_SUFFIXES) {\n    if (pathname.endsWith(suffix)) return true;\n  }\n  return false;\n}\n\nconst BINARY_CONTENT_TYPE_PREFIXES = [\n  'application/pdf',\n  'application/msword',\n  'application/vnd.openxmlformats-officedocument.',\n  'application/vnd.ms-excel',\n  'application/vnd.ms-powerpoint',\n  'application/octet-stream',\n] as const;\n\n/**\n * Post-fetch gate: does this Content-Type header indicate a binary document\n * that our HTML pipeline cannot decode? Returns false for HTML/JSON/plain text\n * and for unknown/missing content-types (the upstream pipeline can still try).\n */\nexport function isBinaryDocumentContentType(contentType: string | null | undefined): boolean {\n  if (!contentType) return false;\n  const lower = contentType.toLowerCase();\n  for (const prefix of BINARY_CONTENT_TYPE_PREFIXES) {\n    if (lower.startsWith(prefix)) return true;\n  }\n  return false;\n}\n"],
-  "mappings": "AAqBA,MAAM,QAAqC;AAAA;AAAA,EAEzC,CAAC,0BAA0B,QAAQ;AAAA,EACnC,CAAC,0BAA0B,QAAQ;AAAA,EACnC,CAAC,0BAA0B,QAAQ;AAAA;AAAA,EAEnC,CAAC,oBAAoB,KAAK;AAAA,EAC1B,CAAC,6BAA6B,KAAK;AAAA,EACnC,CAAC,iCAAiC,IAAI;AAAA,EACtC,CAAC,iCAAiC,IAAI;AAAA,EACtC,CAAC,yBAAyB,OAAO;AAAA,EACjC,CAAC,0BAA0B,MAAM;AAAA,EACjC,CAAC,sBAAsB,MAAM;AAAA,EAC7B,CAAC,4BAA4B,MAAM;AAAA;AAAA,EAEnC,CAAC,6BAA6B,MAAM;AAAA,EACpC,CAAC,aAAa,MAAM;AAAA,EACpB,CAAC,8BAA8B,MAAM;AAAA;AAAA,EAErC,CAAC,gCAAgC,OAAO;AAAA,EACxC,CAAC,wBAAwB,OAAO;AAAA;AAAA,EAEhC,CAAC,+DAA+D,MAAM;AACxE;AAEO,SAAS,oBAAoB,KAAyB;AAC3D,MAAI;AACJ,MAAI;AACF,UAAM,IAAI,IAAI,IAAI,GAAG;AAErB,gBAAY,GAAG,EAAE,QAAQ,GAAG,EAAE,QAAQ;AAAA,EACxC,QAAQ;AACN,gBAAY;AAAA,EACd;AACA,aAAW,CAAC,IAAI,IAAI,KAAK,OAAO;AAC9B,QAAI,GAAG,KAAK,SAAS,EAAG,QAAO;AAAA,EACjC;AACA,SAAO;AACT;AAQA,MAAM,yBAAyB;AAAA,EAC7B;AAAA,EACA;AAAA,EAAQ;AAAA,EACR;AAAA,EAAQ;AAAA,EACR;AAAA,EAAQ;AACV;AAOO,SAAS,cAAc,KAAsB;AAClD,MAAI;AACJ,MAAI;AACF,eAAW,IAAI,IAAI,GAAG,EAAE,SAAS,YAAY;AAAA,EAC/C,QAAQ;AACN,WAAO;AAAA,EACT;AACA,aAAW,UAAU,wBAAwB;AAC3C,QAAI,SAAS,SAAS,MAAM,EAAG,QAAO;AAAA,EACxC;AACA,SAAO;AACT;AAEA,MAAM,+BAA+B;AAAA,EACnC;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF;AAOO,SAAS,4BAA4B,aAAiD;AAC3F,MAAI,CAAC,YAAa,QAAO;AACzB,QAAM,QAAQ,YAAY,YAAY;AACtC,aAAW,UAAU,8BAA8B;AACjD,QAAI,MAAM,WAAW,MAAM,EAAG,QAAO;AAAA,EACvC;AACA,SAAO;AACT;",
+  "mappings": ";AAqBA,IAAM,QAAqC;AAAA;AAAA,EAEzC,CAAC,0BAA0B,QAAQ;AAAA,EACnC,CAAC,0BAA0B,QAAQ;AAAA,EACnC,CAAC,0BAA0B,QAAQ;AAAA;AAAA,EAEnC,CAAC,oBAAoB,KAAK;AAAA,EAC1B,CAAC,6BAA6B,KAAK;AAAA,EACnC,CAAC,iCAAiC,IAAI;AAAA,EACtC,CAAC,iCAAiC,IAAI;AAAA,EACtC,CAAC,yBAAyB,OAAO;AAAA,EACjC,CAAC,0BAA0B,MAAM;AAAA,EACjC,CAAC,sBAAsB,MAAM;AAAA,EAC7B,CAAC,4BAA4B,MAAM;AAAA;AAAA,EAEnC,CAAC,6BAA6B,MAAM;AAAA,EACpC,CAAC,aAAa,MAAM;AAAA,EACpB,CAAC,8BAA8B,MAAM;AAAA;AAAA,EAErC,CAAC,gCAAgC,OAAO;AAAA,EACxC,CAAC,wBAAwB,OAAO;AAAA;AAAA,EAEhC,CAAC,+DAA+D,MAAM;AACxE;AAEO,SAAS,oBAAoB,KAAyB;AAC3D,MAAI;AACJ,MAAI;AACF,UAAM,IAAI,IAAI,IAAI,GAAG;AAErB,gBAAY,GAAG,EAAE,QAAQ,GAAG,EAAE,QAAQ;AAAA,EACxC,QAAQ;AACN,gBAAY;AAAA,EACd;AACA,aAAW,CAAC,IAAI,IAAI,KAAK,OAAO;AAC9B,QAAI,GAAG,KAAK,SAAS,EAAG,QAAO;AAAA,EACjC;AACA,SAAO;AACT;AAQA,IAAM,yBAAyB;AAAA,EAC7B;AAAA,EACA;AAAA,EAAQ;AAAA,EACR;AAAA,EAAQ;AAAA,EACR;AAAA,EAAQ;AACV;AAOO,SAAS,cAAc,KAAsB;AAClD,MAAI;AACJ,MAAI;AACF,eAAW,IAAI,IAAI,GAAG,EAAE,SAAS,YAAY;AAAA,EAC/C,QAAQ;AACN,WAAO;AAAA,EACT;AACA,aAAW,UAAU,wBAAwB;AAC3C,QAAI,SAAS,SAAS,MAAM,EAAG,QAAO;AAAA,EACxC;AACA,SAAO;AACT;AAEA,IAAM,+BAA+B;AAAA,EACnC;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF;AAOO,SAAS,4BAA4B,aAAiD;AAC3F,MAAI,CAAC,YAAa,QAAO;AACzB,QAAM,QAAQ,YAAY,YAAY;AACtC,aAAW,UAAU,8BAA8B;AACjD,QAAI,MAAM,WAAW,MAAM,EAAG,QAAO;AAAA,EACvC;AACA,SAAO;AACT;",
   "names": []
 }

package/dist/src/utils/url-aggregator.js

@@ -1,13 +1,114 @@
-import { CTR_WEIGHTS } from "../config/index.js";
-const WEB_CONSENSUS_THRESHOLD = 3;
-const MIN_BEYOND_TOP10_WEIGHT = 0;
-const BEYOND_TOP10_DECAY = 0.5;
-const BEYOND_TOP10_BASE = 10;
-const DEFAULT_MIN_CONSENSUS_URLS = 5;
-const HIGH_CONSENSUS_THRESHOLD = 4;
-const MAX_ALT_SNIPPETS = 3;
-const MAX_CONSISTENCY_PENALTY = 0.15;
-const CONSISTENCY_STDDEV_SCALE = 5;
+// src/config/index.ts
+import { Logger } from "mcp-use";
+
+// src/version.ts
+import { createRequire } from "module";
+import { fileURLToPath } from "url";
+import { dirname, join } from "path";
+var DEFAULT_PACKAGE_INFO = {
+  version: "3.9.5",
+  name: "mcp-researchpowerpack-http",
+  description: "Research Powerpack MCP Server"
+};
+var packageJson = { ...DEFAULT_PACKAGE_INFO };
+try {
+  if (typeof import.meta.url === "string" && import.meta.url.startsWith("file:")) {
+    const _require = createRequire(import.meta.url);
+    const _dirname = dirname(fileURLToPath(import.meta.url));
+    try {
+      packageJson = _require(join(_dirname, "..", "package.json"));
+    } catch {
+      packageJson = _require(join(_dirname, "..", "..", "package.json"));
+    }
+  }
+} catch {
+}
+var VERSION = packageJson.version;
+var PACKAGE_NAME = packageJson.name;
+var PACKAGE_DESCRIPTION = packageJson.description;
+var USER_AGENT_VERSION = `${PACKAGE_NAME}/${VERSION}`;
+
+// src/config/index.ts
+function safeParseInt(value, defaultVal, min, max) {
+  const logger = Logger.get("config");
+  if (!value) {
+    return defaultVal;
+  }
+  const parsed = parseInt(value, 10);
+  if (isNaN(parsed)) {
+    logger.warn(`Invalid number "${value}", using default ${defaultVal}`);
+    return defaultVal;
+  }
+  if (parsed < min) {
+    logger.warn(`Value ${parsed} below minimum ${min}, clamping to ${min}`);
+    return min;
+  }
+  if (parsed > max) {
+    logger.warn(`Value ${parsed} above maximum ${max}, clamping to ${max}`);
+    return max;
+  }
+  return parsed;
+}
+var CONCURRENCY = {
+  SEARCH: safeParseInt(process.env.CONCURRENCY_SEARCH, 50, 1, 200),
+  SCRAPER: safeParseInt(process.env.CONCURRENCY_SCRAPER, 50, 1, 200),
+  JINA_READER: safeParseInt(process.env.CONCURRENCY_JINA_READER, 50, 1, 200),
+  REDDIT: safeParseInt(process.env.CONCURRENCY_REDDIT, 50, 1, 200),
+  LLM_EXTRACTION: safeParseInt(process.env.LLM_CONCURRENCY, 50, 1, 200),
+  KERNEL: safeParseInt(process.env.CONCURRENCY_KERNEL, 3, 1, 20)
+};
+var CTR_WEIGHTS = {
+  1: 100,
+  2: 60,
+  3: 48.89,
+  4: 33.33,
+  5: 28.89,
+  6: 26.44,
+  7: 24.44,
+  8: 17.78,
+  9: 13.33,
+  10: 12.56
+};
+var cachedLlmExtraction = null;
+function getLlmExtraction() {
+  if (cachedLlmExtraction) return cachedLlmExtraction;
+  const apiKey = process.env.LLM_API_KEY?.trim() || "";
+  const baseUrl = process.env.LLM_BASE_URL?.trim();
+  const model = process.env.LLM_MODEL?.trim();
+  const fallbackModel = process.env.LLM_FALLBACK_MODEL?.trim() || "";
+  if (apiKey && !baseUrl) {
+    throw new Error(
+      "LLM_BASE_URL is required when LLM_API_KEY is set. Set LLM_BASE_URL to your OpenAI-compatible endpoint."
+    );
+  }
+  if (apiKey && !model) {
+    throw new Error(
+      "LLM_MODEL is required when LLM_API_KEY is set."
+    );
+  }
+  cachedLlmExtraction = {
+    API_KEY: apiKey,
+    BASE_URL: baseUrl || "",
+    MODEL: model || "",
+    FALLBACK_MODEL: fallbackModel
+  };
+  return cachedLlmExtraction;
+}
+var LLM_EXTRACTION = new Proxy({}, {
+  get(_target, prop) {
+    return getLlmExtraction()[prop];
+  }
+});
+
+// src/utils/url-aggregator.ts
+var MIN_BEYOND_TOP10_WEIGHT = 0;
+var BEYOND_TOP10_DECAY = 0.5;
+var BEYOND_TOP10_BASE = 10;
+var DEFAULT_MIN_CONSENSUS_URLS = 5;
+var HIGH_CONSENSUS_THRESHOLD = 4;
+var MAX_ALT_SNIPPETS = 3;
+var MAX_CONSISTENCY_PENALTY = 0.15;
+var CONSISTENCY_STDDEV_SCALE = 5;
 function computePositionStats(positions) {
   if (positions.length <= 1) {
     return { mean: positions[0] ?? 0, stdDev: 0, consistencyMultiplier: 1 };
@@ -104,7 +205,7 @@ function calculateWeightedScores(urls, consensusThreshold, totalQueries) {
     consistencyMultiplier: stats.consistencyMultiplier
   }));
 }
-const COVERAGE_TABLE_MAX_ROWS = 20;
+var COVERAGE_TABLE_MAX_ROWS = 20;
 function consistencyLabel(stdDev, frequency) {
   if (frequency <= 1) return "n/a";
   if (stdDev < 1.5) return "high";

package/dist/src/utils/url-aggregator.js.map

@@ -1,7 +1,7 @@
 {
   "version": 3,
-  "sources": ["../../../src/utils/url-aggregator.ts"],
-  "sourcesContent": ["/**\n * URL Aggregator Utility\n * Aggregates search results across multiple queries, calculates CTR-weighted scores,\n * and generates consensus-based rankings.\n */\n\nimport { CTR_WEIGHTS } from '../config/index.js';\nimport type { QuerySearchResult } from '../clients/search.js';\n\n/** Minimum frequency for web search consensus marking */\nconst WEB_CONSENSUS_THRESHOLD = 3 as const;\n\n\n/** Minimum weight assigned to positions beyond top 10 */\nconst MIN_BEYOND_TOP10_WEIGHT = 0 as const;\n\n/** Weight decay per position beyond top 10 */\nconst BEYOND_TOP10_DECAY = 0.5 as const;\n\n/** Base position for beyond-top-10 weight calculation */\nconst BEYOND_TOP10_BASE = 10 as const;\n\n/** Default minimum consensus URLs before lowering threshold (web search) */\nconst DEFAULT_MIN_CONSENSUS_URLS = 5 as const;\n\n/** High consensus frequency threshold for enhanced output labeling */\nconst HIGH_CONSENSUS_THRESHOLD = 4 as const;\n\n/** Maximum number of alternative snippets to retain per URL */\nconst MAX_ALT_SNIPPETS = 3 as const;\n\n/** Consistency penalty cap \u2014 bounds the impact of position variance */\nconst MAX_CONSISTENCY_PENALTY = 0.15 as const;\n\n/** Standard deviation normalizer \u2014 stdDev of 5+ gets full penalty */\nconst CONSISTENCY_STDDEV_SCALE = 5 as const;\n\n/**\n * Aggregated URL data structure\n */\ninterface AggregatedUrl {\n  readonly url: string;\n  title: string;\n  snippet: string;\n  readonly allSnippets: string[];\n  frequency: number;\n  readonly positions: number[];\n  readonly queries: string[];\n  bestPosition: number;\n  totalScore: number;\n}\n\n/**\n * Compute position statistics for consistency scoring\n */\nfunction computePositionStats(positions: number[]): { mean: number; stdDev: number; consistencyMultiplier: number } {\n  if (positions.length <= 1) {\n    return { mean: positions[0] ?? 0, stdDev: 0, consistencyMultiplier: 1.0 };\n  }\n  const mean = positions.reduce((a, b) => a + b, 0) / positions.length;\n  const variance = positions.reduce((sum, p) => sum + (p - mean) ** 2, 0) / (positions.length - 1);\n  const stdDev = Math.sqrt(variance);\n  const consistencyMultiplier = 1.0 - MAX_CONSISTENCY_PENALTY * Math.min(stdDev / CONSISTENCY_STDDEV_SCALE, 1.0);\n  return { mean, stdDev, consistencyMultiplier };\n}\n\n/**\n * Ranked URL with normalized score and enriched signals\n */\ninterface RankedUrl {\n  readonly url: string;\n  readonly title: string;\n  readonly snippet: string;\n  readonly allSnippets: string[];\n  readonly rank: number;\n  readonly score: number;\n  readonly frequency: number;\n  readonly positions: number[];\n  readonly queries: string[];\n  readonly bestPosition: number;\n  readonly isConsensus: boolean;\n  readonly coverageRatio: number;\n  readonly positionStdDev: number;\n  readonly consistencyMultiplier: number;\n}\n\n/**\n * Aggregation result containing all processed data\n */\ninterface AggregationResult {\n  readonly rankedUrls: RankedUrl[];\n  readonly totalUniqueUrls: number;\n  readonly totalQueries: number;\n  readonly frequencyThreshold: number;\n  readonly thresholdNote?: string;\n}\n\n/**\n * Get CTR weight for a position (1-10)\n * Positions beyond 10 get minimal weight\n */\nfunction getCtrWeight(position: number): number {\n  if (position >= 1 && position <= 10) {\n    return CTR_WEIGHTS[position] ?? 0;\n  }\n  // Positions beyond 10 get diminishing returns\n  return Math.max(MIN_BEYOND_TOP10_WEIGHT, BEYOND_TOP10_BASE - (position - BEYOND_TOP10_BASE) * BEYOND_TOP10_DECAY);\n}\n\n/**\n * Aggregate results from multiple searches\n * Flattens all results, deduplicates by URL, and tracks frequency/positions\n */\nfunction aggregateResults(searches: QuerySearchResult[]): Map<string, AggregatedUrl> {\n  const urlMap = new Map<string, AggregatedUrl>();\n\n  for (const search of searches) {\n    for (const result of search.results) {\n      const normalizedUrl = normalizeUrl(result.link);\n      const existing = urlMap.get(normalizedUrl);\n\n      if (existing) {\n        existing.frequency += 1;\n        existing.positions.push(result.position);\n        existing.queries.push(search.query);\n        const prevBest = existing.bestPosition;\n        existing.bestPosition = Math.min(existing.bestPosition, result.position);\n        existing.totalScore += getCtrWeight(result.position);\n        // Collect distinct snippets (up to MAX_ALT_SNIPPETS)\n        if (\n          result.snippet &&\n          existing.allSnippets.length < MAX_ALT_SNIPPETS &&\n          !existing.allSnippets.some(s => s === result.snippet)\n        ) {\n          existing.allSnippets.push(result.snippet);\n        }\n        // Keep best title/snippet (from highest ranking position)\n        if (result.position < prevBest) {\n          existing.title = result.title;\n          existing.snippet = result.snippet;\n        }\n      } else {\n        urlMap.set(normalizedUrl, {\n          url: result.link,\n          title: result.title,\n          snippet: result.snippet,\n          allSnippets: result.snippet ? [result.snippet] : [],\n          frequency: 1,\n          positions: [result.position],\n          queries: [search.query],\n          bestPosition: result.position,\n          totalScore: getCtrWeight(result.position),\n        });\n      }\n    }\n  }\n\n  return urlMap;\n}\n\n/**\n * Normalize URL for deduplication\n * Removes trailing slashes, www prefix, and normalizes protocol\n */\nfunction normalizeUrl(url: string): string {\n  try {\n    const parsed = new URL(url);\n    let host = parsed.hostname.replace(/^www\\./, '');\n    let path = parsed.pathname.replace(/\\/$/, '') || '/';\n    return `${host}${path}${parsed.search}`.toLowerCase();\n  } catch {\n    return url.toLowerCase().replace(/\\/$/, '');\n  }\n}\n\n/**\n * Count URLs meeting a frequency threshold\n */\nfunction countByFrequency(\n  urlMap: Map<string, AggregatedUrl>,\n  minFrequency: number\n): number {\n  let count = 0;\n  for (const url of urlMap.values()) {\n    if (url.frequency >= minFrequency) count++;\n  }\n  return count;\n}\n\n/**\n * Calculate weighted scores with consistency multiplier, normalize to 100.0.\n * Returns ALL URLs sorted by composite score with rank assignments and consensus marking.\n */\nfunction calculateWeightedScores(urls: AggregatedUrl[], consensusThreshold: number, totalQueries: number): RankedUrl[] {\n  if (urls.length === 0) return [];\n\n  // Compute composite scores (base CTR \u00D7 consistency multiplier)\n  const scored = urls.map(url => {\n    const stats = computePositionStats(url.positions);\n    const compositeScore = url.totalScore * stats.consistencyMultiplier;\n    return { url, compositeScore, stats };\n  });\n\n  // Sort by composite score descending\n  scored.sort((a, b) => b.compositeScore - a.compositeScore);\n\n  // Find max for normalization\n  const maxScore = scored[0]!.compositeScore;\n\n  // Map to ranked URLs with all signals\n  return scored.map(({ url, compositeScore, stats }, index) => ({\n    url: url.url,\n    title: url.title,\n    snippet: url.snippet,\n    allSnippets: url.allSnippets,\n    rank: index + 1,\n    score: maxScore > 0 ? (compositeScore / maxScore) * 100 : 0,\n    frequency: url.frequency,\n    positions: url.positions,\n    queries: url.queries,\n    bestPosition: url.bestPosition,\n    isConsensus: url.frequency >= consensusThreshold,\n    coverageRatio: totalQueries > 0 ? url.frequency / totalQueries : 0,\n    positionStdDev: stats.stdDev,\n    consistencyMultiplier: stats.consistencyMultiplier,\n  }));\n}\n\n/** Maximum queries to show in the coverage table before collapsing */\nconst COVERAGE_TABLE_MAX_ROWS = 20 as const;\n\n/**\n * Consistency label based on position standard deviation\n */\nfunction consistencyLabel(stdDev: number, frequency: number): string {\n  if (frequency <= 1) return 'n/a';\n  if (stdDev < 1.5) return 'high';\n  if (stdDev < 3.5) return 'medium';\n  return 'variable';\n}\n\n/**\n * Generate a unified output where every URL appears exactly once.\n * Replaces the old generateEnhancedOutput + per-query section combo.\n */\nexport function generateUnifiedOutput(\n  rankedUrls: RankedUrl[],\n  allQueries: string[],\n  queryResults: QuerySearchResult[],\n  totalUniqueUrls: number,\n  frequencyThreshold: number,\n  thresholdNote?: string,\n  verbose: boolean = false,\n): string {\n  const lines: string[] = [];\n  const consensusCount = rankedUrls.filter(u => u.isConsensus).length;\n\n  // Header\n  lines.push(`## Web Search Results (${allQueries.length} queries, ${totalUniqueUrls} unique URLs)`);\n  lines.push('');\n  if (thresholdNote) {\n    lines.push(`> ${thresholdNote}`);\n    lines.push('');\n  }\n\n  // Ranked URL list \u2014 every URL exactly once.\n  //\n  // Per-row metadata is gated:\n  // - CONSENSUS labels only appear when the effective threshold is >1 (a\n  //   threshold of 1 means *every* row gets the label, so it carries no\n  //   signal). See: docs/code-review/context/02-current-tool-surface.md.\n  // - The Score/Seen/Consistency line is suppressed for rows that were\n  //   seen in exactly one query in a multi-query call (Seen=1/N is common\n  //   and Consistency is always \"n/a\" in that case).\n  // - Verbose mode restores both for callers that explicitly want them.\n  const consensusActive = frequencyThreshold > 1;\n\n  for (const url of rankedUrls) {\n    const consensusTag = consensusActive && url.frequency >= HIGH_CONSENSUS_THRESHOLD\n      ? ' CONSENSUS+++'\n      : consensusActive && url.isConsensus\n        ? ' CONSENSUS'\n        : '';\n    const coveragePct = Math.round(url.coverageRatio * 100);\n    const consistency = consistencyLabel(url.positionStdDev, url.frequency);\n\n    lines.push(`**${url.rank}. [${url.title}](${url.url})**${consensusTag}`);\n\n    const showRowMetadata = verbose\n      || (allQueries.length > 1 && url.frequency > 1)\n      || allQueries.length === 1;\n    if (showRowMetadata) {\n      const parts = [\n        `Score: ${url.score.toFixed(1)}`,\n        `Seen in: ${url.frequency}/${allQueries.length} queries (${coveragePct}%)`,\n        `Best pos: #${url.bestPosition}`,\n      ];\n      if (url.frequency > 1) {\n        parts.push(`Consistency: ${consistency}`);\n      }\n      lines.push(parts.join(' | '));\n    }\n    if (url.queries.length > 1 || verbose) {\n      lines.push(`Queries: ${url.queries.map(q => `\"${q}\"`).join(', ')}`);\n    }\n    lines.push(`> ${url.snippet}`);\n\n    // Alt snippets (if multiple distinct snippets were collected)\n    if (url.allSnippets.length > 1) {\n      const alts = url.allSnippets\n        .filter(s => s !== url.snippet)\n        .slice(0, 3)\n        .map(s => s.length > 100 ? s.slice(0, 97) + '...' : s);\n      if (alts.length > 0) {\n        lines.push(`Alt: ${alts.map(s => `\"${s}\"`).join(' | ')}`);\n      }\n    }\n\n    lines.push('');\n  }\n\n  // Keyword coverage section\n  lines.push('---');\n\n  if (allQueries.length <= COVERAGE_TABLE_MAX_ROWS) {\n    // Full table for \u226420 queries\n    lines.push('### Query Coverage');\n    lines.push('| Query | Results | Top URL | Top Pos |');\n    lines.push('|---------|---------|---------|---------|');\n\n    for (const search of queryResults) {\n      const topResult = search.results[0];\n      let topDomain = '';\n      if (topResult) {\n        try {\n          topDomain = new URL(topResult.link).hostname.replace(/^www\\./, '');\n        } catch {\n          topDomain = topResult.link;\n        }\n      }\n      lines.push(`| \"${search.query}\" | ${search.results.length} | ${topDomain || '\u2014'} | ${topResult ? `#${topResult.position}` : '\u2014'} |`);\n    }\n    lines.push('');\n  } else {\n    // Collapsed summary for >20 queries\n    const goodCount = queryResults.filter(s => s.results.length >= 3).length;\n    lines.push(`### Query Coverage: ${goodCount}/${allQueries.length} queries returned 3+ results`);\n    lines.push('');\n  }\n\n  // Low-yield queries\n  const lowYield = queryResults.filter(s => s.results.length <= 1);\n  if (lowYield.length > 0) {\n    lines.push(`**Low-yield queries** (0-1 results): ${lowYield.map(s => `\\`${s.query}\\``).join(', ')}`);\n    lines.push('');\n  }\n\n  // Related searches (merged and deduplicated)\n  const allRelated = new Set<string>();\n  for (const search of queryResults) {\n    if (search.related) {\n      for (const r of search.related) {\n        allRelated.add(r);\n      }\n    }\n  }\n  if (allRelated.size > 0) {\n    const related = [...allRelated].slice(0, 10);\n    lines.push(`**Related searches:** ${related.map(r => `\\`${r}\\``).join(', ')}`);\n    lines.push('');\n  }\n\n  return lines.join('\\n');\n}\n\n/**\n * Full aggregation pipeline \u2014 returns ALL URLs ranked by CTR score.\n * Determines a consensus threshold (\u22653, \u22652, or \u22651) for labeling, but never\n * drops URLs below the threshold. Every collected URL appears in the output.\n */\nexport function aggregateAndRank(\n  searches: QuerySearchResult[],\n  minConsensusUrls: number = DEFAULT_MIN_CONSENSUS_URLS\n): AggregationResult {\n  const urlMap = aggregateResults(searches);\n  const totalUniqueUrls = urlMap.size;\n  const totalQueries = searches.length;\n\n  // Determine consensus threshold for labeling (not filtering)\n  const thresholds = [3, 2, 1];\n  let usedThreshold = 1;\n  let thresholdNote: string | undefined;\n\n  for (const threshold of thresholds) {\n    const count = countByFrequency(urlMap, threshold);\n    if (count >= minConsensusUrls || threshold === 1) {\n      usedThreshold = threshold;\n      if (threshold < 3) {\n        thresholdNote = `Note: Consensus threshold set to \u2265${threshold} due to result diversity.`;\n      }\n      break;\n    }\n  }\n\n  // Rank ALL URLs, marking consensus based on determined threshold\n  const allUrls = [...urlMap.values()];\n  const rankedUrls = calculateWeightedScores(allUrls, usedThreshold, totalQueries);\n\n  return {\n    rankedUrls,\n    totalUniqueUrls,\n    totalQueries,\n    frequencyThreshold: usedThreshold,\n    thresholdNote,\n  };\n}\n\n"],
-  "mappings": "AAMA,SAAS,mBAAmB;AAI5B,MAAM,0BAA0B;AAIhC,MAAM,0BAA0B;AAGhC,MAAM,qBAAqB;AAG3B,MAAM,oBAAoB;AAG1B,MAAM,6BAA6B;AAGnC,MAAM,2BAA2B;AAGjC,MAAM,mBAAmB;AAGzB,MAAM,0BAA0B;AAGhC,MAAM,2BAA2B;AAoBjC,SAAS,qBAAqB,WAAsF;AAClH,MAAI,UAAU,UAAU,GAAG;AACzB,WAAO,EAAE,MAAM,UAAU,CAAC,KAAK,GAAG,QAAQ,GAAG,uBAAuB,EAAI;AAAA,EAC1E;AACA,QAAM,OAAO,UAAU,OAAO,CAAC,GAAG,MAAM,IAAI,GAAG,CAAC,IAAI,UAAU;AAC9D,QAAM,WAAW,UAAU,OAAO,CAAC,KAAK,MAAM,OAAO,IAAI,SAAS,GAAG,CAAC,KAAK,UAAU,SAAS;AAC9F,QAAM,SAAS,KAAK,KAAK,QAAQ;AACjC,QAAM,wBAAwB,IAAM,0BAA0B,KAAK,IAAI,SAAS,0BAA0B,CAAG;AAC7G,SAAO,EAAE,MAAM,QAAQ,sBAAsB;AAC/C;AAqCA,SAAS,aAAa,UAA0B;AAC9C,MAAI,YAAY,KAAK,YAAY,IAAI;AACnC,WAAO,YAAY,QAAQ,KAAK;AAAA,EAClC;AAEA,SAAO,KAAK,IAAI,yBAAyB,qBAAqB,WAAW,qBAAqB,kBAAkB;AAClH;AAMA,SAAS,iBAAiB,UAA2D;AACnF,QAAM,SAAS,oBAAI,IAA2B;AAE9C,aAAW,UAAU,UAAU;AAC7B,eAAW,UAAU,OAAO,SAAS;AACnC,YAAM,gBAAgB,aAAa,OAAO,IAAI;AAC9C,YAAM,WAAW,OAAO,IAAI,aAAa;AAEzC,UAAI,UAAU;AACZ,iBAAS,aAAa;AACtB,iBAAS,UAAU,KAAK,OAAO,QAAQ;AACvC,iBAAS,QAAQ,KAAK,OAAO,KAAK;AAClC,cAAM,WAAW,SAAS;AAC1B,iBAAS,eAAe,KAAK,IAAI,SAAS,cAAc,OAAO,QAAQ;AACvE,iBAAS,cAAc,aAAa,OAAO,QAAQ;AAEnD,YACE,OAAO,WACP,SAAS,YAAY,SAAS,oBAC9B,CAAC,SAAS,YAAY,KAAK,OAAK,MAAM,OAAO,OAAO,GACpD;AACA,mBAAS,YAAY,KAAK,OAAO,OAAO;AAAA,QAC1C;AAEA,YAAI,OAAO,WAAW,UAAU;AAC9B,mBAAS,QAAQ,OAAO;AACxB,mBAAS,UAAU,OAAO;AAAA,QAC5B;AAAA,MACF,OAAO;AACL,eAAO,IAAI,eAAe;AAAA,UACxB,KAAK,OAAO;AAAA,UACZ,OAAO,OAAO;AAAA,UACd,SAAS,OAAO;AAAA,UAChB,aAAa,OAAO,UAAU,CAAC,OAAO,OAAO,IAAI,CAAC;AAAA,UAClD,WAAW;AAAA,UACX,WAAW,CAAC,OAAO,QAAQ;AAAA,UAC3B,SAAS,CAAC,OAAO,KAAK;AAAA,UACtB,cAAc,OAAO;AAAA,UACrB,YAAY,aAAa,OAAO,QAAQ;AAAA,QAC1C,CAAC;AAAA,MACH;AAAA,IACF;AAAA,EACF;AAEA,SAAO;AACT;AAMA,SAAS,aAAa,KAAqB;AACzC,MAAI;AACF,UAAM,SAAS,IAAI,IAAI,GAAG;AAC1B,QAAI,OAAO,OAAO,SAAS,QAAQ,UAAU,EAAE;AAC/C,QAAI,OAAO,OAAO,SAAS,QAAQ,OAAO,EAAE,KAAK;AACjD,WAAO,GAAG,IAAI,GAAG,IAAI,GAAG,OAAO,MAAM,GAAG,YAAY;AAAA,EACtD,QAAQ;AACN,WAAO,IAAI,YAAY,EAAE,QAAQ,OAAO,EAAE;AAAA,EAC5C;AACF;AAKA,SAAS,iBACP,QACA,cACQ;AACR,MAAI,QAAQ;AACZ,aAAW,OAAO,OAAO,OAAO,GAAG;AACjC,QAAI,IAAI,aAAa,aAAc;AAAA,EACrC;AACA,SAAO;AACT;AAMA,SAAS,wBAAwB,MAAuB,oBAA4B,cAAmC;AACrH,MAAI,KAAK,WAAW,EAAG,QAAO,CAAC;AAG/B,QAAM,SAAS,KAAK,IAAI,SAAO;AAC7B,UAAM,QAAQ,qBAAqB,IAAI,SAAS;AAChD,UAAM,iBAAiB,IAAI,aAAa,MAAM;AAC9C,WAAO,EAAE,KAAK,gBAAgB,MAAM;AAAA,EACtC,CAAC;AAGD,SAAO,KAAK,CAAC,GAAG,MAAM,EAAE,iBAAiB,EAAE,cAAc;AAGzD,QAAM,WAAW,OAAO,CAAC,EAAG;AAG5B,SAAO,OAAO,IAAI,CAAC,EAAE,KAAK,gBAAgB,MAAM,GAAG,WAAW;AAAA,IAC5D,KAAK,IAAI;AAAA,IACT,OAAO,IAAI;AAAA,IACX,SAAS,IAAI;AAAA,IACb,aAAa,IAAI;AAAA,IACjB,MAAM,QAAQ;AAAA,IACd,OAAO,WAAW,IAAK,iBAAiB,WAAY,MAAM;AAAA,IAC1D,WAAW,IAAI;AAAA,IACf,WAAW,IAAI;AAAA,IACf,SAAS,IAAI;AAAA,IACb,cAAc,IAAI;AAAA,IAClB,aAAa,IAAI,aAAa;AAAA,IAC9B,eAAe,eAAe,IAAI,IAAI,YAAY,eAAe;AAAA,IACjE,gBAAgB,MAAM;AAAA,IACtB,uBAAuB,MAAM;AAAA,EAC/B,EAAE;AACJ;AAGA,MAAM,0BAA0B;AAKhC,SAAS,iBAAiB,QAAgB,WAA2B;AACnE,MAAI,aAAa,EAAG,QAAO;AAC3B,MAAI,SAAS,IAAK,QAAO;AACzB,MAAI,SAAS,IAAK,QAAO;AACzB,SAAO;AACT;AAMO,SAAS,sBACd,YACA,YACA,cACA,iBACA,oBACA,eACA,UAAmB,OACX;AACR,QAAM,QAAkB,CAAC;AACzB,QAAM,iBAAiB,WAAW,OAAO,OAAK,EAAE,WAAW,EAAE;AAG7D,QAAM,KAAK,0BAA0B,WAAW,MAAM,aAAa,eAAe,eAAe;AACjG,QAAM,KAAK,EAAE;AACb,MAAI,eAAe;AACjB,UAAM,KAAK,KAAK,aAAa,EAAE;AAC/B,UAAM,KAAK,EAAE;AAAA,EACf;AAYA,QAAM,kBAAkB,qBAAqB;AAE7C,aAAW,OAAO,YAAY;AAC5B,UAAM,eAAe,mBAAmB,IAAI,aAAa,2BACrD,kBACA,mBAAmB,IAAI,cACrB,eACA;AACN,UAAM,cAAc,KAAK,MAAM,IAAI,gBAAgB,GAAG;AACtD,UAAM,cAAc,iBAAiB,IAAI,gBAAgB,IAAI,SAAS;AAEtE,UAAM,KAAK,KAAK,IAAI,IAAI,MAAM,IAAI,KAAK,KAAK,IAAI,GAAG,MAAM,YAAY,EAAE;AAEvE,UAAM,kBAAkB,WAClB,WAAW,SAAS,KAAK,IAAI,YAAY,KAC1C,WAAW,WAAW;AAC3B,QAAI,iBAAiB;AACnB,YAAM,QAAQ;AAAA,QACZ,UAAU,IAAI,MAAM,QAAQ,CAAC,CAAC;AAAA,QAC9B,YAAY,IAAI,SAAS,IAAI,WAAW,MAAM,aAAa,WAAW;AAAA,QACtE,cAAc,IAAI,YAAY;AAAA,MAChC;AACA,UAAI,IAAI,YAAY,GAAG;AACrB,cAAM,KAAK,gBAAgB,WAAW,EAAE;AAAA,MAC1C;AACA,YAAM,KAAK,MAAM,KAAK,KAAK,CAAC;AAAA,IAC9B;AACA,QAAI,IAAI,QAAQ,SAAS,KAAK,SAAS;AACrC,YAAM,KAAK,YAAY,IAAI,QAAQ,IAAI,OAAK,IAAI,CAAC,GAAG,EAAE,KAAK,IAAI,CAAC,EAAE;AAAA,IACpE;AACA,UAAM,KAAK,KAAK,IAAI,OAAO,EAAE;AAG7B,QAAI,IAAI,YAAY,SAAS,GAAG;AAC9B,YAAM,OAAO,IAAI,YACd,OAAO,OAAK,MAAM,IAAI,OAAO,EAC7B,MAAM,GAAG,CAAC,EACV,IAAI,OAAK,EAAE,SAAS,MAAM,EAAE,MAAM,GAAG,EAAE,IAAI,QAAQ,CAAC;AACvD,UAAI,KAAK,SAAS,GAAG;AACnB,cAAM,KAAK,QAAQ,KAAK,IAAI,OAAK,IAAI,CAAC,GAAG,EAAE,KAAK,KAAK,CAAC,EAAE;AAAA,MAC1D;AAAA,IACF;AAEA,UAAM,KAAK,EAAE;AAAA,EACf;AAGA,QAAM,KAAK,KAAK;AAEhB,MAAI,WAAW,UAAU,yBAAyB;AAEhD,UAAM,KAAK,oBAAoB;AAC/B,UAAM,KAAK,yCAAyC;AACpD,UAAM,KAAK,2CAA2C;AAEtD,eAAW,UAAU,cAAc;AACjC,YAAM,YAAY,OAAO,QAAQ,CAAC;AAClC,UAAI,YAAY;AAChB,UAAI,WAAW;AACb,YAAI;AACF,sBAAY,IAAI,IAAI,UAAU,IAAI,EAAE,SAAS,QAAQ,UAAU,EAAE;AAAA,QACnE,QAAQ;AACN,sBAAY,UAAU;AAAA,QACxB;AAAA,MACF;AACA,YAAM,KAAK,MAAM,OAAO,KAAK,OAAO,OAAO,QAAQ,MAAM,MAAM,aAAa,QAAG,MAAM,YAAY,IAAI,UAAU,QAAQ,KAAK,QAAG,IAAI;AAAA,IACrI;AACA,UAAM,KAAK,EAAE;AAAA,EACf,OAAO;AAEL,UAAM,YAAY,aAAa,OAAO,OAAK,EAAE,QAAQ,UAAU,CAAC,EAAE;AAClE,UAAM,KAAK,uBAAuB,SAAS,IAAI,WAAW,MAAM,8BAA8B;AAC9F,UAAM,KAAK,EAAE;AAAA,EACf;AAGA,QAAM,WAAW,aAAa,OAAO,OAAK,EAAE,QAAQ,UAAU,CAAC;AAC/D,MAAI,SAAS,SAAS,GAAG;AACvB,UAAM,KAAK,wCAAwC,SAAS,IAAI,OAAK,KAAK,EAAE,KAAK,IAAI,EAAE,KAAK,IAAI,CAAC,EAAE;AACnG,UAAM,KAAK,EAAE;AAAA,EACf;AAGA,QAAM,aAAa,oBAAI,IAAY;AACnC,aAAW,UAAU,cAAc;AACjC,QAAI,OAAO,SAAS;AAClB,iBAAW,KAAK,OAAO,SAAS;AAC9B,mBAAW,IAAI,CAAC;AAAA,MAClB;AAAA,IACF;AAAA,EACF;AACA,MAAI,WAAW,OAAO,GAAG;AACvB,UAAM,UAAU,CAAC,GAAG,UAAU,EAAE,MAAM,GAAG,EAAE;AAC3C,UAAM,KAAK,yBAAyB,QAAQ,IAAI,OAAK,KAAK,CAAC,IAAI,EAAE,KAAK,IAAI,CAAC,EAAE;AAC7E,UAAM,KAAK,EAAE;AAAA,EACf;AAEA,SAAO,MAAM,KAAK,IAAI;AACxB;AAOO,SAAS,iBACd,UACA,mBAA2B,4BACR;AACnB,QAAM,SAAS,iBAAiB,QAAQ;AACxC,QAAM,kBAAkB,OAAO;AAC/B,QAAM,eAAe,SAAS;AAG9B,QAAM,aAAa,CAAC,GAAG,GAAG,CAAC;AAC3B,MAAI,gBAAgB;AACpB,MAAI;AAEJ,aAAW,aAAa,YAAY;AAClC,UAAM,QAAQ,iBAAiB,QAAQ,SAAS;AAChD,QAAI,SAAS,oBAAoB,cAAc,GAAG;AAChD,sBAAgB;AAChB,UAAI,YAAY,GAAG;AACjB,wBAAgB,0CAAqC,SAAS;AAAA,MAChE;AACA;AAAA,IACF;AAAA,EACF;AAGA,QAAM,UAAU,CAAC,GAAG,OAAO,OAAO,CAAC;AACnC,QAAM,aAAa,wBAAwB,SAAS,eAAe,YAAY;AAE/E,SAAO;AAAA,IACL;AAAA,IACA;AAAA,IACA;AAAA,IACA,oBAAoB;AAAA,IACpB;AAAA,EACF;AACF;",
+  "sources": ["../../../src/config/index.ts", "../../../src/version.ts", "../../../src/utils/url-aggregator.ts"],
+  "sourcesContent": ["/**\n * Consolidated configuration\n * All environment variables, constants, and LLM config in one place\n */\n\nimport { Logger } from 'mcp-use';\n\nimport { VERSION, PACKAGE_NAME, PACKAGE_DESCRIPTION } from '../version.js';\n\n// ============================================================================\n// Safe Integer Parsing Helper\n// ============================================================================\n\n/**\n * Safely parse an integer from environment variable with bounds checking\n */\nfunction safeParseInt(\n  value: string | undefined,\n  defaultVal: number,\n  min: number,\n  max: number\n): number {\n  const logger = Logger.get('config');\n\n  if (!value) {\n    return defaultVal;\n  }\n\n  const parsed = parseInt(value, 10);\n\n  if (isNaN(parsed)) {\n    logger.warn(`Invalid number \"${value}\", using default ${defaultVal}`);\n    return defaultVal;\n  }\n\n  if (parsed < min) {\n    logger.warn(`Value ${parsed} below minimum ${min}, clamping to ${min}`);\n    return min;\n  }\n\n  if (parsed > max) {\n    logger.warn(`Value ${parsed} above maximum ${max}, clamping to ${max}`);\n    return max;\n  }\n\n  return parsed;\n}\n\n\n// ============================================================================\n// Environment Parsing\n// ============================================================================\n\ninterface EnvConfig {\n  SCRAPER_API_KEY: string;\n  SEARCH_API_KEY: string | undefined;\n  REDDIT_CLIENT_ID: string | undefined;\n  REDDIT_CLIENT_SECRET: string | undefined;\n  JINA_API_KEY: string | undefined;\n  KERNEL_API_KEY: string | undefined;\n  KERNEL_PROJECT: string | undefined;\n}\n\nlet cachedEnv: EnvConfig | null = null;\n\nexport function parseEnv(): EnvConfig {\n  if (cachedEnv) return cachedEnv;\n  cachedEnv = {\n    SCRAPER_API_KEY: process.env.SCRAPEDO_API_KEY || '',\n    SEARCH_API_KEY: process.env.SERPER_API_KEY || undefined,\n    REDDIT_CLIENT_ID: process.env.REDDIT_CLIENT_ID || undefined,\n    REDDIT_CLIENT_SECRET: process.env.REDDIT_CLIENT_SECRET || undefined,\n    JINA_API_KEY: process.env.JINA_API_KEY || undefined,\n    KERNEL_API_KEY: process.env.KERNEL_API_KEY || undefined,\n    KERNEL_PROJECT: process.env.KERNEL_PROJECT || undefined,\n  };\n  return cachedEnv;\n}\n\n// ============================================================================\n// MCP Server Configuration\n// ============================================================================\n\nexport const SERVER = {\n  NAME: PACKAGE_NAME,\n  VERSION: VERSION,\n  DESCRIPTION: PACKAGE_DESCRIPTION,\n} as const;\n\n// ============================================================================\n// Capability Detection (which features are available based on ENV)\n// ============================================================================\n\nexport interface Capabilities {\n  reddit: boolean;        // REDDIT_CLIENT_ID + REDDIT_CLIENT_SECRET\n  search: boolean;        // SERPER_API_KEY or JINA_API_KEY\n  serperSearch: boolean;  // SERPER_API_KEY\n  jina: boolean;          // JINA_API_KEY (Reader auth + Search)\n  scraping: boolean;      // SCRAPEDO_API_KEY\n  kernel: boolean;        // KERNEL_API_KEY\n  llmExtraction: boolean; // LLM_API_KEY + LLM_BASE_URL + LLM_MODEL\n}\n\nexport function getCapabilities(): Capabilities {\n  const env = parseEnv();\n  return {\n    reddit: !!(env.REDDIT_CLIENT_ID && env.REDDIT_CLIENT_SECRET),\n    search: !!(env.SEARCH_API_KEY || env.JINA_API_KEY),\n    serperSearch: !!env.SEARCH_API_KEY,\n    jina: !!env.JINA_API_KEY,\n    scraping: !!env.SCRAPER_API_KEY,\n    kernel: !!env.KERNEL_API_KEY,\n    llmExtraction: getLLMConfigStatus().configured,\n  };\n}\n\nexport function getMissingEnvMessage(capability: keyof Capabilities): string {\n  const messages: Record<keyof Capabilities, string> = {\n    reddit: '\u274C **Reddit scraping unavailable.** Set `REDDIT_CLIENT_ID` and `REDDIT_CLIENT_SECRET` to enable threaded Reddit post fetching in `raw-scrape-links` and `smart-scrape-links`.\\n\\n\uD83D\uDC49 Create a Reddit app at: https://www.reddit.com/prefs/apps (select \"script\" type)',\n    search: '\u274C **Search unavailable.** Set `SERPER_API_KEY` or `JINA_API_KEY` to enable `raw-web-search` and `smart-web-search`.\\n\\n\uD83D\uDC49 Serper provides Google SERPs; Jina Search is used as a fallback/search-only provider.',\n    serperSearch: '\u274C **Serper search unavailable.** Set `SERPER_API_KEY` to enable Google-backed primary search.',\n    jina: '\u274C **Jina unavailable.** Set `JINA_API_KEY` to enable Jina Search and authenticated Jina Reader requests.',\n    scraping: '\u26A0\uFE0F **Scrape.do proxy fallback unavailable.** Set `SCRAPEDO_API_KEY` to enable Jina Reader retries through Scrape.do proxy mode.\\n\\n\uD83D\uDC49 Sign up at: https://scrape.do (1,000 free credits)',\n    kernel: '\u274C **Kernel browser rendering unavailable.** Set `KERNEL_API_KEY` to enable optional browser-render fallback for raw/smart scraping.',\n    llmExtraction: '\u26A0\uFE0F **AI extraction disabled.** Set `LLM_API_KEY`, `LLM_BASE_URL`, and `LLM_MODEL` to enable `smart-web-search` and `smart-scrape-links`.\\n\\nUse the raw tools when you need markdown without LLM processing.',\n  };\n  return messages[capability];\n}\n\n// ============================================================================\n// Concurrency Limits\n// ============================================================================\n\nexport const CONCURRENCY = {\n  SEARCH: safeParseInt(process.env.CONCURRENCY_SEARCH, 50, 1, 200),\n  SCRAPER: safeParseInt(process.env.CONCURRENCY_SCRAPER, 50, 1, 200),\n  JINA_READER: safeParseInt(process.env.CONCURRENCY_JINA_READER, 50, 1, 200),\n  REDDIT: safeParseInt(process.env.CONCURRENCY_REDDIT, 50, 1, 200),\n  LLM_EXTRACTION: safeParseInt(process.env.LLM_CONCURRENCY, 50, 1, 200),\n  KERNEL: safeParseInt(process.env.CONCURRENCY_KERNEL, 3, 1, 20),\n} as const;\n\nexport const SCRAPER = {\n  BATCH_SIZE: 30,\n  EXTRACTION_PREFIX: 'Extract from document only \u2014 never hallucinate or add external knowledge.',\n  EXTRACTION_SUFFIX: 'First line = content, not preamble. No confirmation messages.',\n} as const;\n\n// ============================================================================\n// Reddit Configuration\n// ============================================================================\n\nexport const REDDIT = {\n  BATCH_SIZE: 10,\n  MAX_WORDS_PER_POST: 50_000,\n  MAX_WORDS_TOTAL: 500_000,\n  MIN_POSTS: 1,\n  MAX_POSTS: 50,\n  RETRY_COUNT: 5,\n  RETRY_DELAYS: [2000, 4000, 8000, 16000, 32000] as const,\n} as const;\n\n// ============================================================================\n// CTR Weights for URL Ranking (inspired from CTR research)\n// ============================================================================\n\nexport const CTR_WEIGHTS: Record<number, number> = {\n  1: 100.00,\n  2: 60.00,\n  3: 48.89,\n  4: 33.33,\n  5: 28.89,\n  6: 26.44,\n  7: 24.44,\n  8: 17.78,\n  9: 13.33,\n  10: 12.56,\n} as const;\n\n// ============================================================================\n// LLM Configuration\n//\n// Required vars (all must be set together when LLM is enabled):\n//   LLM_API_KEY      \u2014 API key for the OpenAI-compatible endpoint\n//   LLM_BASE_URL     \u2014 endpoint base URL (e.g. https://server.up.railway.app/v1)\n//   LLM_MODEL        \u2014 primary model (e.g. gpt-5.4-mini)\n//\n// Optional:\n//   LLM_FALLBACK_MODEL \u2014 model to use after primary exhausts all retries (e.g. gpt-5.4)\n//   LLM_CONCURRENCY    \u2014 parallel LLM calls (default: 50)\n//\n// Reasoning effort is always 'low' \u2014 not configurable.\n// ============================================================================\n\ninterface LlmExtractionConfig {\n  readonly MODEL: string;\n  readonly FALLBACK_MODEL: string;\n  readonly BASE_URL: string;\n  readonly API_KEY: string;\n}\n\nexport type LLMRequiredEnvVar = 'LLM_API_KEY' | 'LLM_BASE_URL' | 'LLM_MODEL';\n\nexport interface LLMConfigStatus {\n  readonly configured: boolean;\n  readonly apiKeyPresent: boolean;\n  readonly baseUrlPresent: boolean;\n  readonly modelPresent: boolean;\n  readonly missingVars: readonly LLMRequiredEnvVar[];\n  readonly error: string | null;\n}\n\nlet cachedLlmConfigStatus: LLMConfigStatus | null = null;\n\nexport function getLLMConfigStatus(): LLMConfigStatus {\n  if (cachedLlmConfigStatus) return cachedLlmConfigStatus;\n\n  const apiKeyPresent = !!process.env.LLM_API_KEY?.trim();\n  const baseUrlPresent = !!process.env.LLM_BASE_URL?.trim();\n  const modelPresent = !!process.env.LLM_MODEL?.trim();\n  const missingVars: LLMRequiredEnvVar[] = [];\n\n  if (!apiKeyPresent) missingVars.push('LLM_API_KEY');\n  if (!baseUrlPresent) missingVars.push('LLM_BASE_URL');\n  if (!modelPresent) missingVars.push('LLM_MODEL');\n\n  const configured = missingVars.length === 0;\n  cachedLlmConfigStatus = {\n    configured,\n    apiKeyPresent,\n    baseUrlPresent,\n    modelPresent,\n    missingVars,\n    error: configured\n      ? null\n      : `LLM disabled: missing ${missingVars.join(', ')}`,\n  };\n  return cachedLlmConfigStatus;\n}\n\n/**\n * Test-only \u2014 drop every env-derived cache so a test can mutate process.env\n * and re-read fresh values. Covers the LLM config caches AND the parseEnv()\n * cache (which holds SCRAPEDO_API_KEY, SERPER_API_KEY, REDDIT_CLIENT_*,\n * JINA_API_KEY, KERNEL_API_KEY, KERNEL_PROJECT). Tests that scrub non-LLM env vars must also see a clean\n * env on the next parseEnv() call, otherwise getCapabilities() returns\n * stale flags.\n */\nexport function _resetLLMConfigStatusForTests(): void {\n  cachedLlmConfigStatus = null;\n  cachedLlmExtraction = null;\n  cachedEnv = null;\n}\n\nlet cachedLlmExtraction: LlmExtractionConfig | null = null;\n\nfunction getLlmExtraction(): LlmExtractionConfig {\n  if (cachedLlmExtraction) return cachedLlmExtraction;\n\n  const apiKey = process.env.LLM_API_KEY?.trim() || '';\n  const baseUrl = process.env.LLM_BASE_URL?.trim();\n  const model = process.env.LLM_MODEL?.trim();\n  const fallbackModel = process.env.LLM_FALLBACK_MODEL?.trim() || '';\n\n  if (apiKey && !baseUrl) {\n    throw new Error(\n      'LLM_BASE_URL is required when LLM_API_KEY is set. ' +\n      'Set LLM_BASE_URL to your OpenAI-compatible endpoint.',\n    );\n  }\n  if (apiKey && !model) {\n    throw new Error(\n      'LLM_MODEL is required when LLM_API_KEY is set.',\n    );\n  }\n\n  cachedLlmExtraction = {\n    API_KEY: apiKey,\n    BASE_URL: baseUrl || '',\n    MODEL: model || '',\n    FALLBACK_MODEL: fallbackModel,\n  };\n  return cachedLlmExtraction;\n}\n\nexport const LLM_EXTRACTION: LlmExtractionConfig = new Proxy({} as LlmExtractionConfig, {\n  get(_target, prop: string) {\n    return getLlmExtraction()[prop as keyof LlmExtractionConfig];\n  },\n});\n", "/**\n * Version Module - Single Source of Truth\n * \n * This module reads package metadata from package.json at runtime and keeps a\n * fallback copy for environments where package.json cannot be resolved.\n * \n * Usage:\n *   import { VERSION, PACKAGE_NAME } from './version.js';\n */\n\nimport { createRequire } from 'module';\nimport { fileURLToPath } from 'url';\nimport { dirname, join } from 'path';\n\n// Defaults used if package.json cannot be loaded at runtime.\nconst DEFAULT_PACKAGE_INFO = {\n  version: '3.9.5',\n  name: 'mcp-researchpowerpack-http',\n  description: 'Research Powerpack MCP Server',\n} as const;\n\nlet packageJson: { version: string; name: string; description: string } = { ...DEFAULT_PACKAGE_INFO };\n\ntry {\n  if (typeof import.meta.url === 'string' && import.meta.url.startsWith('file:')) {\n    const _require = createRequire(import.meta.url);\n    const _dirname = dirname(fileURLToPath(import.meta.url));\n    try {\n      packageJson = _require(join(_dirname, '..', 'package.json'));\n    } catch {\n      packageJson = _require(join(_dirname, '..', '..', 'package.json'));\n    }\n  }\n} catch {\n  // Keep hardcoded defaults when package.json is unavailable\n}\n\n/**\n * Package version from package.json\n * This is the single source of truth for versioning\n */\nexport const VERSION: string = packageJson.version;\n\n/**\n * Package name from package.json\n */\nexport const PACKAGE_NAME: string = packageJson.name;\n\n/**\n * Package description from package.json\n */\nexport const PACKAGE_DESCRIPTION: string = packageJson.description;\n\n/**\n * Formatted version string for user agents and logging\n * Example: \"mcp-researchpowerpack-http/3.2.0\"\n */\nexport const USER_AGENT_VERSION: string = `${PACKAGE_NAME}/${VERSION}`;\n\n// VERSION_INFO removed - unused, individual exports sufficient\n", "/**\n * URL Aggregator Utility\n * Aggregates search results across multiple queries, calculates CTR-weighted scores,\n * and generates consensus-based rankings.\n */\n\nimport { CTR_WEIGHTS } from '../config/index.js';\nimport type { QuerySearchResult } from '../clients/search.js';\n\n/** Minimum frequency for web search consensus marking */\nconst WEB_CONSENSUS_THRESHOLD = 3 as const;\n\n\n/** Minimum weight assigned to positions beyond top 10 */\nconst MIN_BEYOND_TOP10_WEIGHT = 0 as const;\n\n/** Weight decay per position beyond top 10 */\nconst BEYOND_TOP10_DECAY = 0.5 as const;\n\n/** Base position for beyond-top-10 weight calculation */\nconst BEYOND_TOP10_BASE = 10 as const;\n\n/** Default minimum consensus URLs before lowering threshold (web search) */\nconst DEFAULT_MIN_CONSENSUS_URLS = 5 as const;\n\n/** High consensus frequency threshold for enhanced output labeling */\nconst HIGH_CONSENSUS_THRESHOLD = 4 as const;\n\n/** Maximum number of alternative snippets to retain per URL */\nconst MAX_ALT_SNIPPETS = 3 as const;\n\n/** Consistency penalty cap \u2014 bounds the impact of position variance */\nconst MAX_CONSISTENCY_PENALTY = 0.15 as const;\n\n/** Standard deviation normalizer \u2014 stdDev of 5+ gets full penalty */\nconst CONSISTENCY_STDDEV_SCALE = 5 as const;\n\n/**\n * Aggregated URL data structure\n */\ninterface AggregatedUrl {\n  readonly url: string;\n  title: string;\n  snippet: string;\n  readonly allSnippets: string[];\n  frequency: number;\n  readonly positions: number[];\n  readonly queries: string[];\n  bestPosition: number;\n  totalScore: number;\n}\n\n/**\n * Compute position statistics for consistency scoring\n */\nfunction computePositionStats(positions: number[]): { mean: number; stdDev: number; consistencyMultiplier: number } {\n  if (positions.length <= 1) {\n    return { mean: positions[0] ?? 0, stdDev: 0, consistencyMultiplier: 1.0 };\n  }\n  const mean = positions.reduce((a, b) => a + b, 0) / positions.length;\n  const variance = positions.reduce((sum, p) => sum + (p - mean) ** 2, 0) / (positions.length - 1);\n  const stdDev = Math.sqrt(variance);\n  const consistencyMultiplier = 1.0 - MAX_CONSISTENCY_PENALTY * Math.min(stdDev / CONSISTENCY_STDDEV_SCALE, 1.0);\n  return { mean, stdDev, consistencyMultiplier };\n}\n\n/**\n * Ranked URL with normalized score and enriched signals\n */\ninterface RankedUrl {\n  readonly url: string;\n  readonly title: string;\n  readonly snippet: string;\n  readonly allSnippets: string[];\n  readonly rank: number;\n  readonly score: number;\n  readonly frequency: number;\n  readonly positions: number[];\n  readonly queries: string[];\n  readonly bestPosition: number;\n  readonly isConsensus: boolean;\n  readonly coverageRatio: number;\n  readonly positionStdDev: number;\n  readonly consistencyMultiplier: number;\n}\n\n/**\n * Aggregation result containing all processed data\n */\ninterface AggregationResult {\n  readonly rankedUrls: RankedUrl[];\n  readonly totalUniqueUrls: number;\n  readonly totalQueries: number;\n  readonly frequencyThreshold: number;\n  readonly thresholdNote?: string;\n}\n\n/**\n * Get CTR weight for a position (1-10)\n * Positions beyond 10 get minimal weight\n */\nfunction getCtrWeight(position: number): number {\n  if (position >= 1 && position <= 10) {\n    return CTR_WEIGHTS[position] ?? 0;\n  }\n  // Positions beyond 10 get diminishing returns\n  return Math.max(MIN_BEYOND_TOP10_WEIGHT, BEYOND_TOP10_BASE - (position - BEYOND_TOP10_BASE) * BEYOND_TOP10_DECAY);\n}\n\n/**\n * Aggregate results from multiple searches\n * Flattens all results, deduplicates by URL, and tracks frequency/positions\n */\nfunction aggregateResults(searches: QuerySearchResult[]): Map<string, AggregatedUrl> {\n  const urlMap = new Map<string, AggregatedUrl>();\n\n  for (const search of searches) {\n    for (const result of search.results) {\n      const normalizedUrl = normalizeUrl(result.link);\n      const existing = urlMap.get(normalizedUrl);\n\n      if (existing) {\n        existing.frequency += 1;\n        existing.positions.push(result.position);\n        existing.queries.push(search.query);\n        const prevBest = existing.bestPosition;\n        existing.bestPosition = Math.min(existing.bestPosition, result.position);\n        existing.totalScore += getCtrWeight(result.position);\n        // Collect distinct snippets (up to MAX_ALT_SNIPPETS)\n        if (\n          result.snippet &&\n          existing.allSnippets.length < MAX_ALT_SNIPPETS &&\n          !existing.allSnippets.some(s => s === result.snippet)\n        ) {\n          existing.allSnippets.push(result.snippet);\n        }\n        // Keep best title/snippet (from highest ranking position)\n        if (result.position < prevBest) {\n          existing.title = result.title;\n          existing.snippet = result.snippet;\n        }\n      } else {\n        urlMap.set(normalizedUrl, {\n          url: result.link,\n          title: result.title,\n          snippet: result.snippet,\n          allSnippets: result.snippet ? [result.snippet] : [],\n          frequency: 1,\n          positions: [result.position],\n          queries: [search.query],\n          bestPosition: result.position,\n          totalScore: getCtrWeight(result.position),\n        });\n      }\n    }\n  }\n\n  return urlMap;\n}\n\n/**\n * Normalize URL for deduplication\n * Removes trailing slashes, www prefix, and normalizes protocol\n */\nfunction normalizeUrl(url: string): string {\n  try {\n    const parsed = new URL(url);\n    let host = parsed.hostname.replace(/^www\\./, '');\n    let path = parsed.pathname.replace(/\\/$/, '') || '/';\n    return `${host}${path}${parsed.search}`.toLowerCase();\n  } catch {\n    return url.toLowerCase().replace(/\\/$/, '');\n  }\n}\n\n/**\n * Count URLs meeting a frequency threshold\n */\nfunction countByFrequency(\n  urlMap: Map<string, AggregatedUrl>,\n  minFrequency: number\n): number {\n  let count = 0;\n  for (const url of urlMap.values()) {\n    if (url.frequency >= minFrequency) count++;\n  }\n  return count;\n}\n\n/**\n * Calculate weighted scores with consistency multiplier, normalize to 100.0.\n * Returns ALL URLs sorted by composite score with rank assignments and consensus marking.\n */\nfunction calculateWeightedScores(urls: AggregatedUrl[], consensusThreshold: number, totalQueries: number): RankedUrl[] {\n  if (urls.length === 0) return [];\n\n  // Compute composite scores (base CTR \u00D7 consistency multiplier)\n  const scored = urls.map(url => {\n    const stats = computePositionStats(url.positions);\n    const compositeScore = url.totalScore * stats.consistencyMultiplier;\n    return { url, compositeScore, stats };\n  });\n\n  // Sort by composite score descending\n  scored.sort((a, b) => b.compositeScore - a.compositeScore);\n\n  // Find max for normalization\n  const maxScore = scored[0]!.compositeScore;\n\n  // Map to ranked URLs with all signals\n  return scored.map(({ url, compositeScore, stats }, index) => ({\n    url: url.url,\n    title: url.title,\n    snippet: url.snippet,\n    allSnippets: url.allSnippets,\n    rank: index + 1,\n    score: maxScore > 0 ? (compositeScore / maxScore) * 100 : 0,\n    frequency: url.frequency,\n    positions: url.positions,\n    queries: url.queries,\n    bestPosition: url.bestPosition,\n    isConsensus: url.frequency >= consensusThreshold,\n    coverageRatio: totalQueries > 0 ? url.frequency / totalQueries : 0,\n    positionStdDev: stats.stdDev,\n    consistencyMultiplier: stats.consistencyMultiplier,\n  }));\n}\n\n/** Maximum queries to show in the coverage table before collapsing */\nconst COVERAGE_TABLE_MAX_ROWS = 20 as const;\n\n/**\n * Consistency label based on position standard deviation\n */\nfunction consistencyLabel(stdDev: number, frequency: number): string {\n  if (frequency <= 1) return 'n/a';\n  if (stdDev < 1.5) return 'high';\n  if (stdDev < 3.5) return 'medium';\n  return 'variable';\n}\n\n/**\n * Generate a unified output where every URL appears exactly once.\n * Replaces the old generateEnhancedOutput + per-query section combo.\n */\nexport function generateUnifiedOutput(\n  rankedUrls: RankedUrl[],\n  allQueries: string[],\n  queryResults: QuerySearchResult[],\n  totalUniqueUrls: number,\n  frequencyThreshold: number,\n  thresholdNote?: string,\n  verbose: boolean = false,\n): string {\n  const lines: string[] = [];\n  const consensusCount = rankedUrls.filter(u => u.isConsensus).length;\n\n  // Header\n  lines.push(`## Web Search Results (${allQueries.length} queries, ${totalUniqueUrls} unique URLs)`);\n  lines.push('');\n  if (thresholdNote) {\n    lines.push(`> ${thresholdNote}`);\n    lines.push('');\n  }\n\n  // Ranked URL list \u2014 every URL exactly once.\n  //\n  // Per-row metadata is gated:\n  // - CONSENSUS labels only appear when the effective threshold is >1 (a\n  //   threshold of 1 means *every* row gets the label, so it carries no\n  //   signal). See: docs/code-review/context/02-current-tool-surface.md.\n  // - The Score/Seen/Consistency line is suppressed for rows that were\n  //   seen in exactly one query in a multi-query call (Seen=1/N is common\n  //   and Consistency is always \"n/a\" in that case).\n  // - Verbose mode restores both for callers that explicitly want them.\n  const consensusActive = frequencyThreshold > 1;\n\n  for (const url of rankedUrls) {\n    const consensusTag = consensusActive && url.frequency >= HIGH_CONSENSUS_THRESHOLD\n      ? ' CONSENSUS+++'\n      : consensusActive && url.isConsensus\n        ? ' CONSENSUS'\n        : '';\n    const coveragePct = Math.round(url.coverageRatio * 100);\n    const consistency = consistencyLabel(url.positionStdDev, url.frequency);\n\n    lines.push(`**${url.rank}. [${url.title}](${url.url})**${consensusTag}`);\n\n    const showRowMetadata = verbose\n      || (allQueries.length > 1 && url.frequency > 1)\n      || allQueries.length === 1;\n    if (showRowMetadata) {\n      const parts = [\n        `Score: ${url.score.toFixed(1)}`,\n        `Seen in: ${url.frequency}/${allQueries.length} queries (${coveragePct}%)`,\n        `Best pos: #${url.bestPosition}`,\n      ];\n      if (url.frequency > 1) {\n        parts.push(`Consistency: ${consistency}`);\n      }\n      lines.push(parts.join(' | '));\n    }\n    if (url.queries.length > 1 || verbose) {\n      lines.push(`Queries: ${url.queries.map(q => `\"${q}\"`).join(', ')}`);\n    }\n    lines.push(`> ${url.snippet}`);\n\n    // Alt snippets (if multiple distinct snippets were collected)\n    if (url.allSnippets.length > 1) {\n      const alts = url.allSnippets\n        .filter(s => s !== url.snippet)\n        .slice(0, 3)\n        .map(s => s.length > 100 ? s.slice(0, 97) + '...' : s);\n      if (alts.length > 0) {\n        lines.push(`Alt: ${alts.map(s => `\"${s}\"`).join(' | ')}`);\n      }\n    }\n\n    lines.push('');\n  }\n\n  // Keyword coverage section\n  lines.push('---');\n\n  if (allQueries.length <= COVERAGE_TABLE_MAX_ROWS) {\n    // Full table for \u226420 queries\n    lines.push('### Query Coverage');\n    lines.push('| Query | Results | Top URL | Top Pos |');\n    lines.push('|---------|---------|---------|---------|');\n\n    for (const search of queryResults) {\n      const topResult = search.results[0];\n      let topDomain = '';\n      if (topResult) {\n        try {\n          topDomain = new URL(topResult.link).hostname.replace(/^www\\./, '');\n        } catch {\n          topDomain = topResult.link;\n        }\n      }\n      lines.push(`| \"${search.query}\" | ${search.results.length} | ${topDomain || '\u2014'} | ${topResult ? `#${topResult.position}` : '\u2014'} |`);\n    }\n    lines.push('');\n  } else {\n    // Collapsed summary for >20 queries\n    const goodCount = queryResults.filter(s => s.results.length >= 3).length;\n    lines.push(`### Query Coverage: ${goodCount}/${allQueries.length} queries returned 3+ results`);\n    lines.push('');\n  }\n\n  // Low-yield queries\n  const lowYield = queryResults.filter(s => s.results.length <= 1);\n  if (lowYield.length > 0) {\n    lines.push(`**Low-yield queries** (0-1 results): ${lowYield.map(s => `\\`${s.query}\\``).join(', ')}`);\n    lines.push('');\n  }\n\n  // Related searches (merged and deduplicated)\n  const allRelated = new Set<string>();\n  for (const search of queryResults) {\n    if (search.related) {\n      for (const r of search.related) {\n        allRelated.add(r);\n      }\n    }\n  }\n  if (allRelated.size > 0) {\n    const related = [...allRelated].slice(0, 10);\n    lines.push(`**Related searches:** ${related.map(r => `\\`${r}\\``).join(', ')}`);\n    lines.push('');\n  }\n\n  return lines.join('\\n');\n}\n\n/**\n * Full aggregation pipeline \u2014 returns ALL URLs ranked by CTR score.\n * Determines a consensus threshold (\u22653, \u22652, or \u22651) for labeling, but never\n * drops URLs below the threshold. Every collected URL appears in the output.\n */\nexport function aggregateAndRank(\n  searches: QuerySearchResult[],\n  minConsensusUrls: number = DEFAULT_MIN_CONSENSUS_URLS\n): AggregationResult {\n  const urlMap = aggregateResults(searches);\n  const totalUniqueUrls = urlMap.size;\n  const totalQueries = searches.length;\n\n  // Determine consensus threshold for labeling (not filtering)\n  const thresholds = [3, 2, 1];\n  let usedThreshold = 1;\n  let thresholdNote: string | undefined;\n\n  for (const threshold of thresholds) {\n    const count = countByFrequency(urlMap, threshold);\n    if (count >= minConsensusUrls || threshold === 1) {\n      usedThreshold = threshold;\n      if (threshold < 3) {\n        thresholdNote = `Note: Consensus threshold set to \u2265${threshold} due to result diversity.`;\n      }\n      break;\n    }\n  }\n\n  // Rank ALL URLs, marking consensus based on determined threshold\n  const allUrls = [...urlMap.values()];\n  const rankedUrls = calculateWeightedScores(allUrls, usedThreshold, totalQueries);\n\n  return {\n    rankedUrls,\n    totalUniqueUrls,\n    totalQueries,\n    frequencyThreshold: usedThreshold,\n    thresholdNote,\n  };\n}\n\n"],
+  "mappings": ";AAKA,SAAS,cAAc;;;ACKvB,SAAS,qBAAqB;AAC9B,SAAS,qBAAqB;AAC9B,SAAS,SAAS,YAAY;AAG9B,IAAM,uBAAuB;AAAA,EAC3B,SAAS;AAAA,EACT,MAAM;AAAA,EACN,aAAa;AACf;AAEA,IAAI,cAAsE,EAAE,GAAG,qBAAqB;AAEpG,IAAI;AACF,MAAI,OAAO,YAAY,QAAQ,YAAY,YAAY,IAAI,WAAW,OAAO,GAAG;AAC9E,UAAM,WAAW,cAAc,YAAY,GAAG;AAC9C,UAAM,WAAW,QAAQ,cAAc,YAAY,GAAG,CAAC;AACvD,QAAI;AACF,oBAAc,SAAS,KAAK,UAAU,MAAM,cAAc,CAAC;AAAA,IAC7D,QAAQ;AACN,oBAAc,SAAS,KAAK,UAAU,MAAM,MAAM,cAAc,CAAC;AAAA,IACnE;AAAA,EACF;AACF,QAAQ;AAER;AAMO,IAAM,UAAkB,YAAY;AAKpC,IAAM,eAAuB,YAAY;AAKzC,IAAM,sBAA8B,YAAY;AAMhD,IAAM,qBAA6B,GAAG,YAAY,IAAI,OAAO;;;ADzCpE,SAAS,aACP,OACA,YACA,KACA,KACQ;AACR,QAAM,SAAS,OAAO,IAAI,QAAQ;AAElC,MAAI,CAAC,OAAO;AACV,WAAO;AAAA,EACT;AAEA,QAAM,SAAS,SAAS,OAAO,EAAE;AAEjC,MAAI,MAAM,MAAM,GAAG;AACjB,WAAO,KAAK,mBAAmB,KAAK,oBAAoB,UAAU,EAAE;AACpE,WAAO;AAAA,EACT;AAEA,MAAI,SAAS,KAAK;AAChB,WAAO,KAAK,SAAS,MAAM,kBAAkB,GAAG,iBAAiB,GAAG,EAAE;AACtE,WAAO;AAAA,EACT;AAEA,MAAI,SAAS,KAAK;AAChB,WAAO,KAAK,SAAS,MAAM,kBAAkB,GAAG,iBAAiB,GAAG,EAAE;AACtE,WAAO;AAAA,EACT;AAEA,SAAO;AACT;AAuFO,IAAM,cAAc;AAAA,EACzB,QAAQ,aAAa,QAAQ,IAAI,oBAAoB,IAAI,GAAG,GAAG;AAAA,EAC/D,SAAS,aAAa,QAAQ,IAAI,qBAAqB,IAAI,GAAG,GAAG;AAAA,EACjE,aAAa,aAAa,QAAQ,IAAI,yBAAyB,IAAI,GAAG,GAAG;AAAA,EACzE,QAAQ,aAAa,QAAQ,IAAI,oBAAoB,IAAI,GAAG,GAAG;AAAA,EAC/D,gBAAgB,aAAa,QAAQ,IAAI,iBAAiB,IAAI,GAAG,GAAG;AAAA,EACpE,QAAQ,aAAa,QAAQ,IAAI,oBAAoB,GAAG,GAAG,EAAE;AAC/D;AA0BO,IAAM,cAAsC;AAAA,EACjD,GAAG;AAAA,EACH,GAAG;AAAA,EACH,GAAG;AAAA,EACH,GAAG;AAAA,EACH,GAAG;AAAA,EACH,GAAG;AAAA,EACH,GAAG;AAAA,EACH,GAAG;AAAA,EACH,GAAG;AAAA,EACH,IAAI;AACN;AA6EA,IAAI,sBAAkD;AAEtD,SAAS,mBAAwC;AAC/C,MAAI,oBAAqB,QAAO;AAEhC,QAAM,SAAS,QAAQ,IAAI,aAAa,KAAK,KAAK;AAClD,QAAM,UAAU,QAAQ,IAAI,cAAc,KAAK;AAC/C,QAAM,QAAQ,QAAQ,IAAI,WAAW,KAAK;AAC1C,QAAM,gBAAgB,QAAQ,IAAI,oBAAoB,KAAK,KAAK;AAEhE,MAAI,UAAU,CAAC,SAAS;AACtB,UAAM,IAAI;AAAA,MACR;AAAA,IAEF;AAAA,EACF;AACA,MAAI,UAAU,CAAC,OAAO;AACpB,UAAM,IAAI;AAAA,MACR;AAAA,IACF;AAAA,EACF;AAEA,wBAAsB;AAAA,IACpB,SAAS;AAAA,IACT,UAAU,WAAW;AAAA,IACrB,OAAO,SAAS;AAAA,IAChB,gBAAgB;AAAA,EAClB;AACA,SAAO;AACT;AAEO,IAAM,iBAAsC,IAAI,MAAM,CAAC,GAA0B;AAAA,EACtF,IAAI,SAAS,MAAc;AACzB,WAAO,iBAAiB,EAAE,IAAiC;AAAA,EAC7D;AACF,CAAC;;;AEnRD,IAAM,0BAA0B;AAGhC,IAAM,qBAAqB;AAG3B,IAAM,oBAAoB;AAG1B,IAAM,6BAA6B;AAGnC,IAAM,2BAA2B;AAGjC,IAAM,mBAAmB;AAGzB,IAAM,0BAA0B;AAGhC,IAAM,2BAA2B;AAoBjC,SAAS,qBAAqB,WAAsF;AAClH,MAAI,UAAU,UAAU,GAAG;AACzB,WAAO,EAAE,MAAM,UAAU,CAAC,KAAK,GAAG,QAAQ,GAAG,uBAAuB,EAAI;AAAA,EAC1E;AACA,QAAM,OAAO,UAAU,OAAO,CAAC,GAAG,MAAM,IAAI,GAAG,CAAC,IAAI,UAAU;AAC9D,QAAM,WAAW,UAAU,OAAO,CAAC,KAAK,MAAM,OAAO,IAAI,SAAS,GAAG,CAAC,KAAK,UAAU,SAAS;AAC9F,QAAM,SAAS,KAAK,KAAK,QAAQ;AACjC,QAAM,wBAAwB,IAAM,0BAA0B,KAAK,IAAI,SAAS,0BAA0B,CAAG;AAC7G,SAAO,EAAE,MAAM,QAAQ,sBAAsB;AAC/C;AAqCA,SAAS,aAAa,UAA0B;AAC9C,MAAI,YAAY,KAAK,YAAY,IAAI;AACnC,WAAO,YAAY,QAAQ,KAAK;AAAA,EAClC;AAEA,SAAO,KAAK,IAAI,yBAAyB,qBAAqB,WAAW,qBAAqB,kBAAkB;AAClH;AAMA,SAAS,iBAAiB,UAA2D;AACnF,QAAM,SAAS,oBAAI,IAA2B;AAE9C,aAAW,UAAU,UAAU;AAC7B,eAAW,UAAU,OAAO,SAAS;AACnC,YAAM,gBAAgB,aAAa,OAAO,IAAI;AAC9C,YAAM,WAAW,OAAO,IAAI,aAAa;AAEzC,UAAI,UAAU;AACZ,iBAAS,aAAa;AACtB,iBAAS,UAAU,KAAK,OAAO,QAAQ;AACvC,iBAAS,QAAQ,KAAK,OAAO,KAAK;AAClC,cAAM,WAAW,SAAS;AAC1B,iBAAS,eAAe,KAAK,IAAI,SAAS,cAAc,OAAO,QAAQ;AACvE,iBAAS,cAAc,aAAa,OAAO,QAAQ;AAEnD,YACE,OAAO,WACP,SAAS,YAAY,SAAS,oBAC9B,CAAC,SAAS,YAAY,KAAK,OAAK,MAAM,OAAO,OAAO,GACpD;AACA,mBAAS,YAAY,KAAK,OAAO,OAAO;AAAA,QAC1C;AAEA,YAAI,OAAO,WAAW,UAAU;AAC9B,mBAAS,QAAQ,OAAO;AACxB,mBAAS,UAAU,OAAO;AAAA,QAC5B;AAAA,MACF,OAAO;AACL,eAAO,IAAI,eAAe;AAAA,UACxB,KAAK,OAAO;AAAA,UACZ,OAAO,OAAO;AAAA,UACd,SAAS,OAAO;AAAA,UAChB,aAAa,OAAO,UAAU,CAAC,OAAO,OAAO,IAAI,CAAC;AAAA,UAClD,WAAW;AAAA,UACX,WAAW,CAAC,OAAO,QAAQ;AAAA,UAC3B,SAAS,CAAC,OAAO,KAAK;AAAA,UACtB,cAAc,OAAO;AAAA,UACrB,YAAY,aAAa,OAAO,QAAQ;AAAA,QAC1C,CAAC;AAAA,MACH;AAAA,IACF;AAAA,EACF;AAEA,SAAO;AACT;AAMA,SAAS,aAAa,KAAqB;AACzC,MAAI;AACF,UAAM,SAAS,IAAI,IAAI,GAAG;AAC1B,QAAI,OAAO,OAAO,SAAS,QAAQ,UAAU,EAAE;AAC/C,QAAI,OAAO,OAAO,SAAS,QAAQ,OAAO,EAAE,KAAK;AACjD,WAAO,GAAG,IAAI,GAAG,IAAI,GAAG,OAAO,MAAM,GAAG,YAAY;AAAA,EACtD,QAAQ;AACN,WAAO,IAAI,YAAY,EAAE,QAAQ,OAAO,EAAE;AAAA,EAC5C;AACF;AAKA,SAAS,iBACP,QACA,cACQ;AACR,MAAI,QAAQ;AACZ,aAAW,OAAO,OAAO,OAAO,GAAG;AACjC,QAAI,IAAI,aAAa,aAAc;AAAA,EACrC;AACA,SAAO;AACT;AAMA,SAAS,wBAAwB,MAAuB,oBAA4B,cAAmC;AACrH,MAAI,KAAK,WAAW,EAAG,QAAO,CAAC;AAG/B,QAAM,SAAS,KAAK,IAAI,SAAO;AAC7B,UAAM,QAAQ,qBAAqB,IAAI,SAAS;AAChD,UAAM,iBAAiB,IAAI,aAAa,MAAM;AAC9C,WAAO,EAAE,KAAK,gBAAgB,MAAM;AAAA,EACtC,CAAC;AAGD,SAAO,KAAK,CAAC,GAAG,MAAM,EAAE,iBAAiB,EAAE,cAAc;AAGzD,QAAM,WAAW,OAAO,CAAC,EAAG;AAG5B,SAAO,OAAO,IAAI,CAAC,EAAE,KAAK,gBAAgB,MAAM,GAAG,WAAW;AAAA,IAC5D,KAAK,IAAI;AAAA,IACT,OAAO,IAAI;AAAA,IACX,SAAS,IAAI;AAAA,IACb,aAAa,IAAI;AAAA,IACjB,MAAM,QAAQ;AAAA,IACd,OAAO,WAAW,IAAK,iBAAiB,WAAY,MAAM;AAAA,IAC1D,WAAW,IAAI;AAAA,IACf,WAAW,IAAI;AAAA,IACf,SAAS,IAAI;AAAA,IACb,cAAc,IAAI;AAAA,IAClB,aAAa,IAAI,aAAa;AAAA,IAC9B,eAAe,eAAe,IAAI,IAAI,YAAY,eAAe;AAAA,IACjE,gBAAgB,MAAM;AAAA,IACtB,uBAAuB,MAAM;AAAA,EAC/B,EAAE;AACJ;AAGA,IAAM,0BAA0B;AAKhC,SAAS,iBAAiB,QAAgB,WAA2B;AACnE,MAAI,aAAa,EAAG,QAAO;AAC3B,MAAI,SAAS,IAAK,QAAO;AACzB,MAAI,SAAS,IAAK,QAAO;AACzB,SAAO;AACT;AAMO,SAAS,sBACd,YACA,YACA,cACA,iBACA,oBACA,eACA,UAAmB,OACX;AACR,QAAM,QAAkB,CAAC;AACzB,QAAM,iBAAiB,WAAW,OAAO,OAAK,EAAE,WAAW,EAAE;AAG7D,QAAM,KAAK,0BAA0B,WAAW,MAAM,aAAa,eAAe,eAAe;AACjG,QAAM,KAAK,EAAE;AACb,MAAI,eAAe;AACjB,UAAM,KAAK,KAAK,aAAa,EAAE;AAC/B,UAAM,KAAK,EAAE;AAAA,EACf;AAYA,QAAM,kBAAkB,qBAAqB;AAE7C,aAAW,OAAO,YAAY;AAC5B,UAAM,eAAe,mBAAmB,IAAI,aAAa,2BACrD,kBACA,mBAAmB,IAAI,cACrB,eACA;AACN,UAAM,cAAc,KAAK,MAAM,IAAI,gBAAgB,GAAG;AACtD,UAAM,cAAc,iBAAiB,IAAI,gBAAgB,IAAI,SAAS;AAEtE,UAAM,KAAK,KAAK,IAAI,IAAI,MAAM,IAAI,KAAK,KAAK,IAAI,GAAG,MAAM,YAAY,EAAE;AAEvE,UAAM,kBAAkB,WAClB,WAAW,SAAS,KAAK,IAAI,YAAY,KAC1C,WAAW,WAAW;AAC3B,QAAI,iBAAiB;AACnB,YAAM,QAAQ;AAAA,QACZ,UAAU,IAAI,MAAM,QAAQ,CAAC,CAAC;AAAA,QAC9B,YAAY,IAAI,SAAS,IAAI,WAAW,MAAM,aAAa,WAAW;AAAA,QACtE,cAAc,IAAI,YAAY;AAAA,MAChC;AACA,UAAI,IAAI,YAAY,GAAG;AACrB,cAAM,KAAK,gBAAgB,WAAW,EAAE;AAAA,MAC1C;AACA,YAAM,KAAK,MAAM,KAAK,KAAK,CAAC;AAAA,IAC9B;AACA,QAAI,IAAI,QAAQ,SAAS,KAAK,SAAS;AACrC,YAAM,KAAK,YAAY,IAAI,QAAQ,IAAI,OAAK,IAAI,CAAC,GAAG,EAAE,KAAK,IAAI,CAAC,EAAE;AAAA,IACpE;AACA,UAAM,KAAK,KAAK,IAAI,OAAO,EAAE;AAG7B,QAAI,IAAI,YAAY,SAAS,GAAG;AAC9B,YAAM,OAAO,IAAI,YACd,OAAO,OAAK,MAAM,IAAI,OAAO,EAC7B,MAAM,GAAG,CAAC,EACV,IAAI,OAAK,EAAE,SAAS,MAAM,EAAE,MAAM,GAAG,EAAE,IAAI,QAAQ,CAAC;AACvD,UAAI,KAAK,SAAS,GAAG;AACnB,cAAM,KAAK,QAAQ,KAAK,IAAI,OAAK,IAAI,CAAC,GAAG,EAAE,KAAK,KAAK,CAAC,EAAE;AAAA,MAC1D;AAAA,IACF;AAEA,UAAM,KAAK,EAAE;AAAA,EACf;AAGA,QAAM,KAAK,KAAK;AAEhB,MAAI,WAAW,UAAU,yBAAyB;AAEhD,UAAM,KAAK,oBAAoB;AAC/B,UAAM,KAAK,yCAAyC;AACpD,UAAM,KAAK,2CAA2C;AAEtD,eAAW,UAAU,cAAc;AACjC,YAAM,YAAY,OAAO,QAAQ,CAAC;AAClC,UAAI,YAAY;AAChB,UAAI,WAAW;AACb,YAAI;AACF,sBAAY,IAAI,IAAI,UAAU,IAAI,EAAE,SAAS,QAAQ,UAAU,EAAE;AAAA,QACnE,QAAQ;AACN,sBAAY,UAAU;AAAA,QACxB;AAAA,MACF;AACA,YAAM,KAAK,MAAM,OAAO,KAAK,OAAO,OAAO,QAAQ,MAAM,MAAM,aAAa,QAAG,MAAM,YAAY,IAAI,UAAU,QAAQ,KAAK,QAAG,IAAI;AAAA,IACrI;AACA,UAAM,KAAK,EAAE;AAAA,EACf,OAAO;AAEL,UAAM,YAAY,aAAa,OAAO,OAAK,EAAE,QAAQ,UAAU,CAAC,EAAE;AAClE,UAAM,KAAK,uBAAuB,SAAS,IAAI,WAAW,MAAM,8BAA8B;AAC9F,UAAM,KAAK,EAAE;AAAA,EACf;AAGA,QAAM,WAAW,aAAa,OAAO,OAAK,EAAE,QAAQ,UAAU,CAAC;AAC/D,MAAI,SAAS,SAAS,GAAG;AACvB,UAAM,KAAK,wCAAwC,SAAS,IAAI,OAAK,KAAK,EAAE,KAAK,IAAI,EAAE,KAAK,IAAI,CAAC,EAAE;AACnG,UAAM,KAAK,EAAE;AAAA,EACf;AAGA,QAAM,aAAa,oBAAI,IAAY;AACnC,aAAW,UAAU,cAAc;AACjC,QAAI,OAAO,SAAS;AAClB,iBAAW,KAAK,OAAO,SAAS;AAC9B,mBAAW,IAAI,CAAC;AAAA,MAClB;AAAA,IACF;AAAA,EACF;AACA,MAAI,WAAW,OAAO,GAAG;AACvB,UAAM,UAAU,CAAC,GAAG,UAAU,EAAE,MAAM,GAAG,EAAE;AAC3C,UAAM,KAAK,yBAAyB,QAAQ,IAAI,OAAK,KAAK,CAAC,IAAI,EAAE,KAAK,IAAI,CAAC,EAAE;AAC7E,UAAM,KAAK,EAAE;AAAA,EACf;AAEA,SAAO,MAAM,KAAK,IAAI;AACxB;AAOO,SAAS,iBACd,UACA,mBAA2B,4BACR;AACnB,QAAM,SAAS,iBAAiB,QAAQ;AACxC,QAAM,kBAAkB,OAAO;AAC/B,QAAM,eAAe,SAAS;AAG9B,QAAM,aAAa,CAAC,GAAG,GAAG,CAAC;AAC3B,MAAI,gBAAgB;AACpB,MAAI;AAEJ,aAAW,aAAa,YAAY;AAClC,UAAM,QAAQ,iBAAiB,QAAQ,SAAS;AAChD,QAAI,SAAS,oBAAoB,cAAc,GAAG;AAChD,sBAAgB;AAChB,UAAI,YAAY,GAAG;AACjB,wBAAgB,0CAAqC,SAAS;AAAA,MAChE;AACA;AAAA,IACF;AAAA,EACF;AAGA,QAAM,UAAU,CAAC,GAAG,OAAO,OAAO,CAAC;AACnC,QAAM,aAAa,wBAAwB,SAAS,eAAe,YAAY;AAE/E,SAAO;AAAA,IACL;AAAA,IACA;AAAA,IACA;AAAA,IACA,oBAAoB;AAAA,IACpB;AAAA,EACF;AACF;",
   "names": []
 }

package/dist/src/version.js

@@ -1,12 +1,13 @@
+// src/version.ts
 import { createRequire } from "module";
 import { fileURLToPath } from "url";
 import { dirname, join } from "path";
-const DEFAULT_PACKAGE_INFO = {
+var DEFAULT_PACKAGE_INFO = {
   version: "3.9.5",
   name: "mcp-researchpowerpack-http",
   description: "Research Powerpack MCP Server"
 };
-let packageJson = { ...DEFAULT_PACKAGE_INFO };
+var packageJson = { ...DEFAULT_PACKAGE_INFO };
 try {
   if (typeof import.meta.url === "string" && import.meta.url.startsWith("file:")) {
     const _require = createRequire(import.meta.url);
@@ -19,10 +20,10 @@ try {
   }
 } catch {
 }
-const VERSION = packageJson.version;
-const PACKAGE_NAME = packageJson.name;
-const PACKAGE_DESCRIPTION = packageJson.description;
-const USER_AGENT_VERSION = `${PACKAGE_NAME}/${VERSION}`;
+var VERSION = packageJson.version;
+var PACKAGE_NAME = packageJson.name;
+var PACKAGE_DESCRIPTION = packageJson.description;
+var USER_AGENT_VERSION = `${PACKAGE_NAME}/${VERSION}`;
 export {
   PACKAGE_DESCRIPTION,
   PACKAGE_NAME,

package/dist/src/version.js.map

@@ -2,6 +2,6 @@
   "version": 3,
   "sources": ["../../src/version.ts"],
   "sourcesContent": ["/**\n * Version Module - Single Source of Truth\n * \n * This module reads package metadata from package.json at runtime and keeps a\n * fallback copy for environments where package.json cannot be resolved.\n * \n * Usage:\n *   import { VERSION, PACKAGE_NAME } from './version.js';\n */\n\nimport { createRequire } from 'module';\nimport { fileURLToPath } from 'url';\nimport { dirname, join } from 'path';\n\n// Defaults used if package.json cannot be loaded at runtime.\nconst DEFAULT_PACKAGE_INFO = {\n  version: '3.9.5',\n  name: 'mcp-researchpowerpack-http',\n  description: 'Research Powerpack MCP Server',\n} as const;\n\nlet packageJson: { version: string; name: string; description: string } = { ...DEFAULT_PACKAGE_INFO };\n\ntry {\n  if (typeof import.meta.url === 'string' && import.meta.url.startsWith('file:')) {\n    const _require = createRequire(import.meta.url);\n    const _dirname = dirname(fileURLToPath(import.meta.url));\n    try {\n      packageJson = _require(join(_dirname, '..', 'package.json'));\n    } catch {\n      packageJson = _require(join(_dirname, '..', '..', 'package.json'));\n    }\n  }\n} catch {\n  // Keep hardcoded defaults when package.json is unavailable\n}\n\n/**\n * Package version from package.json\n * This is the single source of truth for versioning\n */\nexport const VERSION: string = packageJson.version;\n\n/**\n * Package name from package.json\n */\nexport const PACKAGE_NAME: string = packageJson.name;\n\n/**\n * Package description from package.json\n */\nexport const PACKAGE_DESCRIPTION: string = packageJson.description;\n\n/**\n * Formatted version string for user agents and logging\n * Example: \"mcp-researchpowerpack-http/3.2.0\"\n */\nexport const USER_AGENT_VERSION: string = `${PACKAGE_NAME}/${VERSION}`;\n\n// VERSION_INFO removed - unused, individual exports sufficient\n"],
-  "mappings": "AAUA,SAAS,qBAAqB;AAC9B,SAAS,qBAAqB;AAC9B,SAAS,SAAS,YAAY;AAG9B,MAAM,uBAAuB;AAAA,EAC3B,SAAS;AAAA,EACT,MAAM;AAAA,EACN,aAAa;AACf;AAEA,IAAI,cAAsE,EAAE,GAAG,qBAAqB;AAEpG,IAAI;AACF,MAAI,OAAO,YAAY,QAAQ,YAAY,YAAY,IAAI,WAAW,OAAO,GAAG;AAC9E,UAAM,WAAW,cAAc,YAAY,GAAG;AAC9C,UAAM,WAAW,QAAQ,cAAc,YAAY,GAAG,CAAC;AACvD,QAAI;AACF,oBAAc,SAAS,KAAK,UAAU,MAAM,cAAc,CAAC;AAAA,IAC7D,QAAQ;AACN,oBAAc,SAAS,KAAK,UAAU,MAAM,MAAM,cAAc,CAAC;AAAA,IACnE;AAAA,EACF;AACF,QAAQ;AAER;AAMO,MAAM,UAAkB,YAAY;AAKpC,MAAM,eAAuB,YAAY;AAKzC,MAAM,sBAA8B,YAAY;AAMhD,MAAM,qBAA6B,GAAG,YAAY,IAAI,OAAO;",
+  "mappings": ";AAUA,SAAS,qBAAqB;AAC9B,SAAS,qBAAqB;AAC9B,SAAS,SAAS,YAAY;AAG9B,IAAM,uBAAuB;AAAA,EAC3B,SAAS;AAAA,EACT,MAAM;AAAA,EACN,aAAa;AACf;AAEA,IAAI,cAAsE,EAAE,GAAG,qBAAqB;AAEpG,IAAI;AACF,MAAI,OAAO,YAAY,QAAQ,YAAY,YAAY,IAAI,WAAW,OAAO,GAAG;AAC9E,UAAM,WAAW,cAAc,YAAY,GAAG;AAC9C,UAAM,WAAW,QAAQ,cAAc,YAAY,GAAG,CAAC;AACvD,QAAI;AACF,oBAAc,SAAS,KAAK,UAAU,MAAM,cAAc,CAAC;AAAA,IAC7D,QAAQ;AACN,oBAAc,SAAS,KAAK,UAAU,MAAM,MAAM,cAAc,CAAC;AAAA,IACnE;AAAA,EACF;AACF,QAAQ;AAER;AAMO,IAAM,UAAkB,YAAY;AAKpC,IAAM,eAAuB,YAAY;AAKzC,IAAM,sBAA8B,YAAY;AAMhD,IAAM,qBAA6B,GAAG,YAAY,IAAI,OAAO;",
   "names": []
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mcp-researchpowerpack",
-  "version": "7.0.11",
+  "version": "7.0.13",
   "description": "HTTP-first MCP research server: start-research plus raw/smart search and scrape tools — built on mcp-use.",
   "type": "module",
   "main": "dist/index.js",
@@ -39,7 +39,7 @@
     "url": "https://github.com/yigitkonur/mcp-researchpowerpack/issues"
   },
   "dependencies": {
-    "@mcp-use/cli": "2.20.0",
+    "@mcp-use/cli": "3.1.2",
     "@mozilla/readability": "^0.6.0",
     "@onkernel/sdk": "^0.52.0",
     "@types/jsdom": "^28.0.1",
@@ -47,7 +47,7 @@
     "@types/turndown": "^5.0.6",
     "effect": "^3.21.2",
     "jsdom": "^29.0.2",
-    "mcp-use": "1.22.0",
+    "mcp-use": "1.26.0",
     "openai": "^4.77.0",
     "p-map": "^7.0.3",
     "react": "^19.0.0",