@claude-flow/cli 3.10.23 → 3.10.25

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@claude-flow/cli",
-  "version": "3.10.23",
+  "version": "3.10.25",
   "type": "module",
   "description": "Ruflo CLI - Enterprise AI agent orchestration with 60+ specialized agents, swarm coordination, MCP server, self-learning hooks, and vector memory for Claude Code",
   "main": "dist/src/index.js",

package/scripts/benchmark-cross-repo.mjs

@@ -0,0 +1,206 @@
+#!/usr/bin/env node
+// benchmark-cross-repo.mjs — cross-repo generalisation proof (ADR-084).
+//
+// Pretrains on a DIFFERENT repo's history (default: ruvnet/agentdb) and runs
+// labelled queries about THAT repo's work. If the retrieval system genuinely
+// generalises (vs overfits to the ruflo corpus it was tuned on), nDCG@3
+// should stay near 0.96.
+//
+// Usage:
+//   1. cd /tmp/agentdb-bench
+//   2. REPO_ROOT=/tmp/agentdb-bench GH_REPO=ruvnet/agentdb \
+//        node /path/to/pretrain-from-github.mjs    # writes /tmp/agentdb-bench/.claude-flow/neural/
+//   3. cd /tmp/agentdb-bench && node /path/to/benchmark-cross-repo.mjs
+//
+// Or use the wrapper at the bottom that does both steps with --repo=<name>.
+
+import { writeFileSync, mkdirSync } from 'node:fs';
+import { fileURLToPath } from 'node:url';
+import { dirname, join, resolve } from 'node:path';
+import { performance } from 'node:perf_hooks';
+
+const SCRIPT_DIR = dirname(fileURLToPath(import.meta.url));
+const CLI_ROOT = resolve(SCRIPT_DIR, '..');
+const RUFLO_ROOT = resolve(SCRIPT_DIR, '../../../..');
+const RUNS_DIR = join(RUFLO_ROOT, 'docs', 'benchmarks', 'runs');
+
+// Labelled queries about agentdb's actual history (commits + issues
+// observed via `git log` + `gh issue list` on ruvnet/agentdb).
+// Each query targets a specific piece of work; labels are case-insensitive
+// substring matches against the pattern name.
+const QUERIES_AGENTDB = [
+  { q: 'SQL injection fix REINDEX',
+    labels: ['sql injection', 'reindex'] },
+  { q: 'insecure PRNG random IDs',
+    labels: ['insecure prng', 'random ids', 'random id', 'crypto', 'uuid'] },
+  { q: 'JSON parse crash hardening',
+    labels: ['json.parse', 'json parse', 'crash hardening'] },
+  { q: 'protobuf critical CVE patch',
+    labels: ['protobufjs', 'protobuf', 'cve'] },
+  { q: 'spawnSync vs execSync security',
+    labels: ['spawnsync', 'execsync'] },
+  { q: 'better-sqlite3 inline schemas',
+    labels: ['better-sqlite3', 'inline schema', 'schemas'] },
+  { q: 'Claude Code marketplace plugins',
+    labels: ['marketplace', 'plugin', 'claude code'] },
+  { q: 'README rewrite ruflo style',
+    labels: ['readme', 'rewrite'] },
+  { q: 'WASM browser fs.readFileSync blocking',
+    labels: ['wasm', 'browser', 'fs.readfilesync', 'readfilesync', 'agentdb/wasm'] },
+  { q: 'AgentDB missing schema files',
+    labels: ['missing schema', 'mising schema', 'schema files', 'not working'] },
+];
+
+// Labelled queries about agentic-flow's actual history.
+const QUERIES_AGENTIC_FLOW = [
+  { q: 'CWE-78 shell injection fix',
+    labels: ['cwe-78', 'shell injection', 'execsync', 'safe-exec'] },
+  { q: 'SSRF hardcoded key NaN panic security',
+    labels: ['ssrf', 'hardcoded key', 'nan-panic', 'nan panic'] },
+  { q: 'WebSocket QUIC transport fallback',
+    labels: ['websocket', 'quic', 'transport'] },
+  { q: 'agentdb submodule bump',
+    labels: ['agentdb submodule', 'bump agentdb', 'submodule'] },
+  { q: 'sql.js prepared statement leak',
+    labels: ['sql.js', 'prepared statement', 'leak'] },
+  { q: 'WASM ADR-071 integration',
+    labels: ['wasm', 'adr-071', 'wasm-integration'] },
+  { q: 'patch CVEs transitive overrides',
+    labels: ['cve', 'patch', 'transitive', 'overrides'] },
+  { q: 'GraphDatabaseAdapter delete API',
+    labels: ['graphdatabaseadapter', 'delete api', 'reflexionmemory', 'deleteepisode'] },
+  { q: 'release alpha publish version',
+    labels: ['release', 'publish', 'alpha', 'version'] },
+  { q: 'extract packages to dedicated repo',
+    labels: ['extract', 'refactor', 'dedicated repo', 'submodule', 'agentdb'] },
+];
+
+// Auto-pick query set based on GH_REPO env var.
+const QUERY_SETS = {
+  'ruvnet/agentdb': QUERIES_AGENTDB,
+  'ruvnet/agentic-flow': QUERIES_AGENTIC_FLOW,
+};
+const QUERIES = QUERY_SETS[process.env.GH_REPO] || QUERIES_AGENTDB;
+const TOP_K = 5;
+
+function isRelevant(name, labels) {
+  if (!name || !labels?.length) return false;
+  const lower = String(name).toLowerCase();
+  return labels.some((s) => lower.includes(s.toLowerCase()));
+}
+
+function ndcgAtK(rankedRel, k) {
+  const arr = rankedRel.slice(0, k);
+  const dcg = arr.reduce((acc, rel, i) => acc + (rel ? 1 / Math.log2(i + 2) : 0), 0);
+  const numRelevant = arr.filter(Boolean).length;
+  if (numRelevant === 0) return 0;
+  let idcg = 0;
+  for (let i = 0; i < numRelevant; i++) idcg += 1 / Math.log2(i + 2);
+  return idcg > 0 ? dcg / idcg : 0;
+}
+
+async function evalConfig(tool, params) {
+  let top1 = 0, top3 = 0, ranks = [], ndcg3sum = 0, ndcg5sum = 0, p3sum = 0;
+  const t0 = performance.now();
+  const perQuery = [];
+  for (const { q, labels } of QUERIES) {
+    const r = await tool.handler({ action: 'search', query: q, mode: 'hybrid', limit: TOP_K, ...params });
+    const matches = (r.results || []).slice(0, TOP_K);
+    const rel = matches.map((m) => isRelevant(m?.name, labels));
+    if (rel[0]) top1++;
+    if (rel.slice(0, 3).some(Boolean)) top3++;
+    const firstRank = rel.findIndex(Boolean);
+    if (firstRank >= 0) ranks.push(firstRank + 1);
+    ndcg3sum += ndcgAtK(rel, 3);
+    ndcg5sum += ndcgAtK(rel, 5);
+    p3sum += rel.slice(0, 3).filter(Boolean).length / 3;
+    perQuery.push({ q, top1: rel[0] || false, topK: matches.map((m, i) => ({ name: m.name?.slice(0, 90), relevant: rel[i], score: (m.score ?? m.similarity)?.toFixed?.(3) })) });
+  }
+  const elapsed = performance.now() - t0;
+  return {
+    top1HitRate: top1 / QUERIES.length,
+    top3HitRate: top3 / QUERIES.length,
+    mrr3: ranks.reduce((s, r) => s + 1 / r, 0) / QUERIES.length,
+    precision3: p3sum / QUERIES.length,
+    ndcg3: ndcg3sum / QUERIES.length,
+    ndcg5: ndcg5sum / QUERIES.length,
+    avgLatencyMs: elapsed / QUERIES.length,
+    perQuery,
+  };
+}
+
+async function main() {
+  const neural = await import(join(CLI_ROOT, 'dist/src/mcp-tools/neural-tools.js'));
+  const intel = await import(join(CLI_ROOT, 'dist/src/memory/intelligence.js'));
+  const stats = await intel.getUnifiedLearningStats();
+  const storeSize = stats.neuralPatterns.patternCount;
+
+  if (storeSize === 0) {
+    console.error('No patterns in neural store. Run pretrain-from-github.mjs first (with REPO_ROOT + GH_REPO env).');
+    process.exit(2);
+  }
+
+  const tool = neural.neuralTools.find((t) => t.name === 'neural_patterns');
+  if (!tool) throw new Error('neural_patterns tool not found');
+
+  console.log(`# Cross-repo generalisation benchmark (ADR-084)`);
+  console.log(`Cwd: ${process.cwd()}`);
+  console.log(`Store size: ${storeSize} patterns`);
+  console.log(`Queries: ${QUERIES.length} (labelled set for ${process.env.GH_REPO || 'default agentdb'})`);
+  console.log('');
+
+  console.log('--- Hybrid path (defaults) ---');
+  const hybrid = await evalConfig(tool, { rerank: false });
+  console.log(`  top-1: ${(hybrid.top1HitRate * 100).toFixed(0)}%  top-3: ${(hybrid.top3HitRate * 100).toFixed(0)}%  MRR@3: ${hybrid.mrr3.toFixed(3)}  P3: ${hybrid.precision3.toFixed(3)}  nDCG@3: ${hybrid.ndcg3.toFixed(3)}  nDCG@5: ${hybrid.ndcg5.toFixed(3)}  ${hybrid.avgLatencyMs.toFixed(0)}ms`);
+
+  console.log('--- Rerank path (defaults) ---');
+  const rerank = await evalConfig(tool, { rerank: true });
+  console.log(`  top-1: ${(rerank.top1HitRate * 100).toFixed(0)}%  top-3: ${(rerank.top3HitRate * 100).toFixed(0)}%  MRR@3: ${rerank.mrr3.toFixed(3)}  P3: ${rerank.precision3.toFixed(3)}  nDCG@3: ${rerank.ndcg3.toFixed(3)}  nDCG@5: ${rerank.ndcg5.toFixed(3)}  ${rerank.avgLatencyMs.toFixed(0)}ms`);
+
+  console.log('\n--- Per-query rerank top-3 ---');
+  for (const q of rerank.perQuery) {
+    console.log(`Q: "${q.q}" (top-1: ${q.top1 ? '✓' : '✗'})`);
+    for (const m of q.topK.slice(0, 3)) {
+      console.log(`   ${m.relevant ? '★' : ' '} [${m.score ?? '—'}] ${m.name}`);
+    }
+  }
+
+  const summary = {
+    runAt: new Date().toISOString(),
+    benchmark: 'cross-repo-generalisation',
+    repo: process.env.GH_REPO || 'unknown',
+    cwd: process.cwd(),
+    storeSize,
+    queries: QUERIES.length,
+    hybrid: {
+      top1HitRate: hybrid.top1HitRate,
+      top3HitRate: hybrid.top3HitRate,
+      mrr3: hybrid.mrr3,
+      precision3: hybrid.precision3,
+      ndcg3: hybrid.ndcg3,
+      ndcg5: hybrid.ndcg5,
+      avgLatencyMs: hybrid.avgLatencyMs,
+    },
+    rerank: {
+      top1HitRate: rerank.top1HitRate,
+      top3HitRate: rerank.top3HitRate,
+      mrr3: rerank.mrr3,
+      precision3: rerank.precision3,
+      ndcg3: rerank.ndcg3,
+      ndcg5: rerank.ndcg5,
+      avgLatencyMs: rerank.avgLatencyMs,
+    },
+    perQueryRerank: rerank.perQuery,
+  };
+
+  mkdirSync(RUNS_DIR, { recursive: true });
+  const stamp = summary.runAt.replace(/[:.]/g, '-');
+  const repoSlug = (process.env.GH_REPO || 'unknown').replace('/', '-');
+  writeFileSync(join(RUNS_DIR, `cross-repo-${repoSlug}-${stamp}.json`), JSON.stringify(summary, null, 2));
+  writeFileSync(join(RUNS_DIR, `cross-repo-${repoSlug}-latest.json`), JSON.stringify(summary, null, 2));
+  console.log(`\nWrote ${join(RUNS_DIR, `cross-repo-${repoSlug}-${stamp}.json`)}`);
+
+  process.exit(0);
+}
+
+main().catch((err) => { console.error(err); process.exit(1); });

package/scripts/pretrain-from-github.mjs

@@ -32,8 +32,14 @@ import { performance } from 'node:perf_hooks';
 
 const SCRIPT_DIR = dirname(fileURLToPath(import.meta.url));
 const CLI_ROOT = resolve(SCRIPT_DIR, '..');
-const REPO_ROOT = resolve(SCRIPT_DIR, '../../../..');
-const RUNS_DIR = join(REPO_ROOT, 'docs', 'benchmarks', 'runs');
+const RUFLO_ROOT = resolve(SCRIPT_DIR, '../../../..');
+const RUNS_DIR = join(RUFLO_ROOT, 'docs', 'benchmarks', 'runs');
+
+// ADR-084 cross-repo support — REPO_ROOT and GH_REPO env-overridable so the
+// same script can pretrain on agentdb / agentic-flow / any other repo for
+// generalisation testing. Defaults preserve ruflo behaviour.
+const REPO_ROOT = process.env.REPO_ROOT ? resolve(process.env.REPO_ROOT) : RUFLO_ROOT;
+const GH_REPO   = process.env.GH_REPO   || 'ruvnet/ruflo';
 
 const COMMITS = Number(process.env.COMMITS) || 50;
 const ISSUES  = Number(process.env.ISSUES)  || 30;
@@ -156,7 +162,7 @@ function harvestIssues(n) {
   if (SOURCE === 'git') return [];
   try {
     const raw = execSync(
-      `gh issue list --repo ruvnet/ruflo --state all --limit ${n} --json number,title,body,state,closedAt 2>/dev/null`,
+      `gh issue list --repo ${GH_REPO} --state all --limit ${n} --json number,title,body,state,closedAt 2>/dev/null`,
       { cwd: REPO_ROOT, encoding: 'utf-8', maxBuffer: 64 * 1024 * 1024 },
     );
     const items = JSON.parse(raw);

package/scripts/run-beir-bge.mjs

@@ -0,0 +1,292 @@
+#!/usr/bin/env node
+// run-beir-bge.mjs — NFCorpus retrieval using real BGE-base-en-v1.5
+// embeddings (bypasses neural-tools' broken agentic-flow path that hash-fell-
+// back on darwin-arm64 without sharp/libvips).
+//
+// Stores doc embeddings in a single .f32 binary file (3633 * 768 * 4 = ~11MB)
+// for fast bench iteration without re-embedding.
+//
+// Usage:
+//   cd /tmp/beir-nfcorpus
+//   node /path/to/scripts/run-beir-bge.mjs                          # full ingest + bench
+//   SKIP_INGEST=1 node /path/to/scripts/run-beir-bge.mjs            # reuse cached embeds
+//   BGE_MODEL=Xenova/bge-small-en-v1.5 node ...                     # faster, lower-quality
+//   BGE_MODEL=Xenova/bge-large-en-v1.5 node ...                     # slower, higher-quality
+//
+// ADR-086.
+
+import { readFileSync, writeFileSync, existsSync, mkdirSync } from 'node:fs';
+import { fileURLToPath } from 'node:url';
+import { dirname, join, resolve } from 'node:path';
+import { performance } from 'node:perf_hooks';
+
+const SCRIPT_DIR = dirname(fileURLToPath(import.meta.url));
+const CLI_ROOT = resolve(SCRIPT_DIR, '..');
+const RUFLO_ROOT = resolve(SCRIPT_DIR, '../../../..');
+const RUNS_DIR = join(RUFLO_ROOT, 'docs', 'benchmarks', 'runs');
+
+const DATA_DIR = process.env.BEIR_DATA_DIR || '/tmp/beir-nfcorpus/nfcorpus';
+const CACHE_DIR = join('/tmp/beir-nfcorpus', 'bge-cache');
+const BGE_MODEL = process.env.BGE_MODEL || 'Xenova/bge-base-en-v1.5';
+const SKIP_INGEST = process.env.SKIP_INGEST === '1';
+const MAX_QUERIES = Number(process.env.MAX_QUERIES) || 0;
+
+// Published BEIR baselines per dataset (nDCG@10).
+// Sources: Thakur et al. 2021 (BEIR paper), BGE paper (BAAI 2024),
+// SPLADE++ paper, papers-with-code BEIR leaderboards.
+const BASELINES_BY_DATASET = {
+  nfcorpus: {
+    'BM25 (Lucene)':        0.325,
+    'DocT5query':           0.328,
+    'TAS-B':                0.319,
+    'GenQ':                 0.319,
+    'ColBERT':              0.305,
+    'Contriever':           0.328,
+    'GTR-XL':               0.343,
+    'SPLADE++':             0.347,
+    'BGE-large-v1.5 (pub)': 0.380,
+    'SBERT msmarco':        0.272,
+  },
+  scifact: {
+    'BM25 (Lucene)':        0.679,
+    'DocT5query':           0.675,
+    'TAS-B':                0.643,
+    'GenQ':                 0.644,
+    'ColBERT':              0.671,
+    'Contriever':           0.677,
+    'GTR-XL':               0.662,
+    'SPLADE++':             0.704,
+    'BGE-large-v1.5 (pub)': 0.722,
+    'SBERT msmarco':        0.555,
+  },
+};
+
+// Auto-detect dataset from DATA_DIR path; default to nfcorpus baselines.
+function detectDataset(path) {
+  const p = path.toLowerCase();
+  for (const ds of Object.keys(BASELINES_BY_DATASET)) {
+    if (p.includes(ds)) return ds;
+  }
+  return 'nfcorpus';
+}
+
+// ---------------------------------------------------------------------------
+// nDCG (graded)
+// ---------------------------------------------------------------------------
+
+function dcg(rels, k) {
+  let s = 0;
+  for (let i = 0; i < Math.min(rels.length, k); i++) {
+    s += (Math.pow(2, rels[i]) - 1) / Math.log2(i + 2);
+  }
+  return s;
+}
+function ndcg(retrieved, qrels, k) {
+  const rels = retrieved.slice(0, k).map((id) => qrels.get(id) ?? 0);
+  const ideal = [...qrels.values()].sort((a, b) => b - a).slice(0, k);
+  const idcg = dcg(ideal, k);
+  return idcg > 0 ? dcg(rels, k) / idcg : 0;
+}
+function mrr(retrieved, qrels, k) {
+  for (let i = 0; i < Math.min(retrieved.length, k); i++) {
+    if ((qrels.get(retrieved[i]) ?? 0) > 0) return 1 / (i + 1);
+  }
+  return 0;
+}
+function recall(retrieved, qrels, k) {
+  const tot = [...qrels.values()].filter((v) => v > 0).length;
+  if (tot === 0) return 0;
+  let h = 0;
+  for (let i = 0; i < Math.min(retrieved.length, k); i++) {
+    if ((qrels.get(retrieved[i]) ?? 0) > 0) h++;
+  }
+  return h / tot;
+}
+
+function cosine(a, b) {
+  let s = 0;
+  for (let i = 0; i < a.length; i++) s += a[i] * b[i];
+  return s; // already L2-normalised in BGE
+}
+
+// ---------------------------------------------------------------------------
+// Loaders
+// ---------------------------------------------------------------------------
+
+function loadJsonl(path) {
+  return readFileSync(path, 'utf-8').split('\n').filter(Boolean).map((l) => JSON.parse(l));
+}
+function loadQrels(path) {
+  const qrels = new Map();
+  const lines = readFileSync(path, 'utf-8').split('\n');
+  for (let i = 1; i < lines.length; i++) {
+    if (!lines[i].trim()) continue;
+    const [qid, did, score] = lines[i].split('\t');
+    if (!qrels.has(qid)) qrels.set(qid, new Map());
+    qrels.get(qid).set(did, Number(score));
+  }
+  return qrels;
+}
+
+// ---------------------------------------------------------------------------
+// Embedding cache I/O — single .f32 file + .ids file
+// ---------------------------------------------------------------------------
+
+function cachePath(suffix) {
+  const safe = BGE_MODEL.replace(/\//g, '_');
+  return join(CACHE_DIR, `${safe}.${suffix}`);
+}
+
+function saveEmbeddings(ids, embeddings, dim) {
+  mkdirSync(CACHE_DIR, { recursive: true });
+  // ids: one per line
+  writeFileSync(cachePath('ids'), ids.join('\n') + '\n');
+  // embeddings: concatenated Float32Array
+  const buf = new Float32Array(ids.length * dim);
+  for (let i = 0; i < ids.length; i++) buf.set(embeddings[i], i * dim);
+  writeFileSync(cachePath('f32'), Buffer.from(buf.buffer));
+}
+
+function loadEmbeddings(dim) {
+  if (!existsSync(cachePath('ids')) || !existsSync(cachePath('f32'))) return null;
+  const ids = readFileSync(cachePath('ids'), 'utf-8').split('\n').filter(Boolean);
+  const raw = readFileSync(cachePath('f32'));
+  const buf = new Float32Array(raw.buffer, raw.byteOffset, raw.length / 4);
+  const embeds = [];
+  for (let i = 0; i < ids.length; i++) embeds.push(buf.slice(i * dim, (i + 1) * dim));
+  return { ids, embeds };
+}
+
+// ---------------------------------------------------------------------------
+// Main
+// ---------------------------------------------------------------------------
+
+async function main() {
+  const dataset = detectDataset(DATA_DIR);
+  const BASELINES_NDCG10 = BASELINES_BY_DATASET[dataset];
+  console.log(`# BEIR ${dataset} — BGE embedder (ADR-085/086)`);
+  console.log(`Model: ${BGE_MODEL}`);
+  console.log(`Data:  ${DATA_DIR}`);
+
+  const corpus = loadJsonl(join(DATA_DIR, 'corpus.jsonl'));
+  const queries = loadJsonl(join(DATA_DIR, 'queries.jsonl'));
+  const qrels = loadQrels(join(DATA_DIR, 'qrels/test.tsv'));
+  console.log(`Corpus: ${corpus.length} docs · Queries: ${queries.length} · Test qrels: ${qrels.size}`);
+
+  const bge = await import(join(CLI_ROOT, 'dist/src/memory/bge-embedder.js'));
+  const emb = await bge.getBgeEmbedder(BGE_MODEL);
+  if (!emb) {
+    console.error('BGE failed to load:', bge.getBgeStatus());
+    process.exit(1);
+  }
+  const dim = emb.dim();
+  console.log(`BGE loaded (dim=${dim})`);
+
+  // Ingest or load cached doc embeddings.
+  let docIds, docEmbeds;
+  const cached = SKIP_INGEST ? loadEmbeddings(dim) : null;
+  if (cached) {
+    docIds = cached.ids;
+    docEmbeds = cached.embeds;
+    console.log(`Loaded ${docIds.length} cached doc embeddings from ${cachePath('f32')}`);
+  } else {
+    console.log(`\nEmbedding ${corpus.length} docs (this is the slow step, ~20 min on M-series CPU)...`);
+    docIds = corpus.map((d) => d._id);
+    docEmbeds = new Array(corpus.length);
+    const tIngest = performance.now();
+    for (let i = 0; i < corpus.length; i++) {
+      const d = corpus[i];
+      const text = `${d.title || ''}\n${d.text || ''}`.slice(0, 4096);
+      docEmbeds[i] = await emb.embed(text);
+      if ((i + 1) % 100 === 0) {
+        const elapsed = (performance.now() - tIngest) / 1000;
+        const rate = (i + 1) / elapsed;
+        const eta = (corpus.length - i - 1) / rate;
+        console.log(`  ${i + 1}/${corpus.length}  ${rate.toFixed(1)} docs/s  ETA ${eta.toFixed(0)}s`);
+      }
+    }
+    console.log(`Embedded ${corpus.length} docs in ${((performance.now() - tIngest) / 1000).toFixed(0)}s`);
+    saveEmbeddings(docIds, docEmbeds, dim);
+    console.log(`Cached to ${cachePath('f32')}`);
+  }
+
+  // Run retrieval per query.
+  const queriesById = new Map(queries.map((q) => [q._id, q.text]));
+  const evalQueryIds = MAX_QUERIES
+    ? [...qrels.keys()].slice(0, MAX_QUERIES)
+    : [...qrels.keys()];
+
+  console.log(`\nRunning ${evalQueryIds.length} queries...`);
+  let nSum = 0, mSum = 0, r10Sum = 0, r100Sum = 0, n = 0;
+  const tQ = performance.now();
+  for (const qid of evalQueryIds) {
+    const qtext = queriesById.get(qid);
+    if (!qtext) continue;
+    // BGE recommends prepending "Represent this sentence for searching relevant passages:" to queries
+    // but only for v1.5 *en-icl* variants. v1.5 (non-icl) works without it.
+    const qEmb = await emb.embed(qtext);
+    // Score every doc, sort.
+    const scores = new Array(docEmbeds.length);
+    for (let i = 0; i < docEmbeds.length; i++) scores[i] = { id: docIds[i], score: cosine(qEmb, docEmbeds[i]) };
+    scores.sort((a, b) => b.score - a.score);
+    const top100 = scores.slice(0, 100).map((s) => s.id);
+
+    const qmap = qrels.get(qid);
+    nSum += ndcg(top100, qmap, 10);
+    mSum += mrr(top100, qmap, 10);
+    r10Sum += recall(top100, qmap, 10);
+    r100Sum += recall(top100, qmap, 100);
+    n++;
+    if (n % 50 === 0) {
+      const elapsed = (performance.now() - tQ) / 1000;
+      console.log(`  ${n}/${evalQueryIds.length} in ${elapsed.toFixed(0)}s  running nDCG@10=${(nSum / n).toFixed(4)}`);
+    }
+  }
+  const queryMs = performance.now() - tQ;
+
+  const ndcg10 = nSum / n;
+  const mrr10 = mSum / n;
+  const recall10 = r10Sum / n;
+  const recall100 = r100Sum / n;
+
+  console.log(`\n=== Results (N=${n}, BGE-base-en-v1.5) ===`);
+  console.log(`  nDCG@10:    ${ndcg10.toFixed(4)}`);
+  console.log(`  MRR@10:     ${mrr10.toFixed(4)}`);
+  console.log(`  Recall@10:  ${recall10.toFixed(4)}`);
+  console.log(`  Recall@100: ${recall100.toFixed(4)}`);
+  console.log(`  Avg query latency: ${(queryMs / n).toFixed(0)}ms`);
+
+  console.log(`\n=== vs published NFCorpus baselines (nDCG@10) ===`);
+  const ourLabel = `ruflo + ${BGE_MODEL.replace('Xenova/', '')}`;
+  const ranking = [
+    ...Object.entries(BASELINES_NDCG10).map(([name, score]) => ({ name, score, ours: false })),
+    { name: ourLabel, score: ndcg10, ours: true },
+  ].sort((a, b) => b.score - a.score);
+  for (const r of ranking) {
+    console.log(`  ${r.score.toFixed(3)}  ${r.name}${r.ours ? ' ← us' : ''}`);
+  }
+  const ourRank = ranking.findIndex((r) => r.ours) + 1;
+  console.log(`\n  Our rank: ${ourRank} / ${ranking.length}`);
+
+  const summary = {
+    runAt: new Date().toISOString(),
+    benchmark: `beir-${dataset}-bge`,
+    dataset,
+    model: BGE_MODEL,
+    queries: n,
+    corpusSize: corpus.length,
+    metrics: { ndcg10, mrr10, recall10, recall100, avgQueryLatencyMs: queryMs / n },
+    baselines: BASELINES_NDCG10,
+    ourRank,
+    leaderboardLength: ranking.length,
+  };
+  mkdirSync(RUNS_DIR, { recursive: true });
+  const stamp = summary.runAt.replace(/[:.]/g, '-');
+  const safe = BGE_MODEL.replace(/\//g, '_');
+  writeFileSync(join(RUNS_DIR, `beir-${dataset}-bge-${safe}-${stamp}.json`), JSON.stringify(summary, null, 2));
+  writeFileSync(join(RUNS_DIR, `beir-${dataset}-bge-latest.json`), JSON.stringify(summary, null, 2));
+  console.log(`\nWrote ${join(RUNS_DIR, `beir-${dataset}-bge-${safe}-${stamp}.json`)}`);
+  process.exit(0);
+}
+
+main().catch((err) => { console.error(err); process.exit(1); });