ruvnet-brain 4.0.1 → 4.0.2

package/scripts/ingest-meeting.mjs

@@ -0,0 +1,115 @@
+#!/usr/bin/env node
+// ingest-meeting.mjs — turn a Read.ai meeting transcript into brain passages (the ruv-meetings store).
+//
+// WHY (2026-07-17, Stuart): rUv's community calls are where he explains HOW HE THINKS — the
+// doctrine behind the repos (meta-proxy load balancing, flywheel economics, oracle memory,
+// harness-vs-model philosophy). The repos hold the WHAT; the calls hold the WHY. This makes them
+// citable: search_ruvnet returns "ruv-meetings/<date>-<slug>" passages alongside source code.
+//
+// PRIVACY — this store is PRIVATE BY CONSTRUCTION and that is load-bearing:
+//   • kb/ruv-meetings.* is gitignored (never committed to the public repo)
+//   • "ruv-meetings" is listed in kb/PRIVATE-STORES.json, so scripts/build-bundle.mjs excludes it
+//     from every publishable bundle (same fail-closed fence as the cognitum stores)
+//   Participants of a community call did not consent to public redistribution. Local brain only.
+//
+// Output contract mirrors ingest-gists.mjs: kb/ruv-meetings.passages.jsonl (+ .meta.json), then:
+//   node kb/forge-big.mjs both --dir kb --name ruv-meetings
+//
+// Usage:
+//   node scripts/ingest-meeting.mjs --file "docs/Ruv_advice/Ruv hackerspace Transcript.txt" \
+//        --title "Ruv hackerspace" --date 2026-07-16 [--append]
+//   --append keeps existing passages (multi-meeting store grows over time); default rebuilds.
+
+import fs from 'node:fs';
+import path from 'node:path';
+import { fileURLToPath } from 'node:url';
+
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
+const KB = path.join(path.dirname(__dirname), 'kb');
+const OUT_PASSAGES = path.join(KB, 'ruv-meetings.passages.jsonl');
+const OUT_META = path.join(KB, 'ruv-meetings.meta.json');
+
+const args = process.argv.slice(2);
+const get = (k) => { const i = args.indexOf('--' + k); return i >= 0 ? args[i + 1] : null; };
+const file = get('file');
+const title = get('title') || 'community call';
+const date = get('date') || 'unknown-date';
+const APPEND = args.includes('--append');
+if (!file || !fs.existsSync(file)) { console.error('need --file <transcript.txt>'); process.exit(2); }
+
+const raw = fs.readFileSync(file, 'utf8');
+
+// Read.ai format: blocks of "H:MM(:SS) - Speaker Name\n<text>\n". Parse into turns, then pack turns
+// into passages of ~2,200 chars, never splitting inside a turn — a speaker's thought stays whole
+// (the same clean-thought law the explainer's line breaks follow).
+const turnRe = /^(\d{1,2}:\d{2}(?::\d{2})?) - (.+)$/;
+const lines = raw.split('\n');
+const turns = [];
+let cur = null;
+for (const line of lines) {
+  const m = line.match(turnRe);
+  if (m) {
+    if (cur && cur.text.trim()) turns.push(cur);
+    cur = { ts: m[1], speaker: m[2].trim(), text: '' };
+  } else if (cur) {
+    cur.text += (cur.text ? '\n' : '') + line;
+  }
+}
+if (cur && cur.text.trim()) turns.push(cur);
+
+const slug = title.toLowerCase().replace(/[^a-z0-9]+/g, '-').replace(/^-|-$/g, '');
+const HEADER = `SOURCE: rUv community call — "${title}" (${date}), Read.ai transcript.\n`
+  + `STATUS: spoken, informal — rUv's live thinking and practice. Treat as INTENT and doctrine;\n`
+  + `verify shipped behavior against repo source before asserting a capability exists.\n`;
+
+const TARGET = 2200;
+const passages = [];
+let buf = [];
+let bufLen = 0;
+let firstTs = null;
+const flush = () => {
+  if (!buf.length) return;
+  const seg = String(passages.length).padStart(4, '0');
+  passages.push({
+    id: String(passages.length),
+    // UNIQUE path per segment (verified fix, 2026-07-19). A shared path made forge-ask.mjs doc-collapse
+    // crush every segment of a meeting into ONE ~12k window — the 317-passages→4-paths collapse that
+    // made meeting recall unretrievable (facts were IN the corpus but never surfaced; meeting recall
+    // 3/28). One path = one segment, the same granular convention ingest-gists.mjs uses (gist-id/file#N).
+    // With this + the reader's transcript-store BM25 candidates, meeting recall went 3/28 → 25/28. See
+    // docs/adr/0025.
+    path: `ruv-meetings/${date}/${slug}-seg-${seg}`,
+    title: `${title} — ${firstTs}`,
+    text: `${HEADER}segment starting ${firstTs} —\n\n` + buf.join('\n\n'),
+    meeting: slug, date, ts: firstTs,
+  });
+  buf = []; bufLen = 0; firstTs = null;
+};
+for (const t of turns) {
+  const block = `[${t.ts}] ${t.speaker}:\n${t.text.trim()}`;
+  if (bufLen + block.length > TARGET && buf.length) flush();
+  if (!buf.length) firstTs = t.ts;
+  buf.push(block); bufLen += block.length;
+}
+flush();
+
+let existing = [];
+if (APPEND && fs.existsSync(OUT_PASSAGES)) {
+  existing = fs.readFileSync(OUT_PASSAGES, 'utf8').split('\n').filter(Boolean).map((l) => JSON.parse(l));
+  // Re-ingesting the same meeting replaces its old passages instead of duplicating them.
+  existing = existing.filter((p) => p.meeting !== slug);
+}
+const all = [...existing, ...passages].map((p, i) => ({ ...p, id: String(i) }));
+fs.writeFileSync(OUT_PASSAGES, all.map((p) => JSON.stringify(p)).join('\n') + '\n');
+fs.writeFileSync(OUT_META, JSON.stringify({
+  builder: 'ingest-meeting',
+  builtUtc: new Date().toISOString(),
+  private: true,
+  fence: 'kb/PRIVATE-STORES.json + .gitignore — never bundle, never commit',
+  meetings: [...new Set(all.map((p) => `${p.date} ${p.meeting}`))],
+  passages: all.length,
+}, null, 2) + '\n');
+
+console.log(`ingest-meeting: ${turns.length} turns → ${passages.length} passages for "${title}" (${date})`);
+console.log(`total store: ${all.length} passages → ${path.relative(process.cwd(), OUT_PASSAGES)}`);
+console.log(`next: node kb/forge-big.mjs both --dir kb --name ruv-meetings   (embed → ruv-meetings.big.rvf)`);

package/scripts/ingest-repo.mjs

@@ -0,0 +1,79 @@
+#!/usr/bin/env node
+// ingest-repo.mjs — pull ANY github.com/ruvnet/<name> repo into the brain ON DEMAND.
+//
+//   clone (shallow) -> transactional incremental bge-768 refresh -> symbol index
+//
+// After it finishes the new repo is searchable immediately: search_ruvnet / forge-ask-all discover
+// every <name>.rvf in the kb dir at query time, so no server restart is needed. Use this when a
+// project needs a RuvNet repo the brain doesn't cover yet — load it first, don't guess about it.
+//
+//   node scripts/ingest-repo.mjs --name rvm
+//
+// For full capability-confidence on the new repo (so it's never wrongly doubted), also build its
+// prose primer afterwards:
+//   node scripts/build-primer.mjs --name <name> --variant big
+//   node scripts/build-concepts.mjs && node kb/forge-big.mjs both --dir kb --name concepts
+//
+// Portable: paths resolve relative to this file; model cache via KB_MODEL_CACHE (or auto-download).
+import path from 'node:path';
+import fs from 'node:fs';
+import { execFileSync } from 'node:child_process';
+import { fileURLToPath } from 'node:url';
+import { FULL_HINTS, KEEP_DIRS } from './full-hints.mjs';
+
+const ROOT = path.resolve(path.dirname(fileURLToPath(import.meta.url)), '..');
+const arg = (f, d) => { const i = process.argv.indexOf(f); return i >= 0 && process.argv[i + 1] ? process.argv[i + 1] : d; };
+const NAME = arg('--name');
+// --org lets us ingest ecosystem repos that live in a rUv COLLABORATOR org (e.g. the QE fleet at
+// proffesor-for-testing/agentic-qe), not just github.com/ruvnet/*. Defaults to ruvnet.
+const ORG = arg('--org', 'ruvnet');
+const SOURCE = arg('--source');
+if (!NAME) {
+  console.error('Usage: node scripts/ingest-repo.mjs --name <repo> [--org <github-org>] [--source <local-checkout>]');
+  process.exit(2);
+}
+
+const KB = path.join(ROOT, 'kb');
+const CLONES = path.join(ROOT, 'clones');
+const kb = NAME.toLowerCase();
+const dir = SOURCE ? path.resolve(SOURCE) : path.join(CLONES, NAME);
+const env = { ...process.env };
+const run = (cmd, args, opts) => execFileSync(cmd, args, { stdio: 'inherit', ...opts });
+
+fs.mkdirSync(CLONES, { recursive: true });
+if (SOURCE) {
+  if (!fs.existsSync(path.join(dir, '.git'))) {
+    console.error(`[source] ${dir} is not a git checkout`);
+    process.exit(2);
+  }
+  console.log(`[source] local candidate ${dir}`);
+} else if (!fs.existsSync(path.join(dir, '.git'))) {
+  console.log(`[clone] ${ORG}/${NAME}`);
+  run('git', ['clone', '--depth', '1', `https://github.com/${ORG}/${NAME}`, dir]);
+} else {
+  console.log(`[update] ${ORG}/${NAME}`);
+  run('git', ['-C', dir, 'fetch', '--depth', '1', 'origin']);
+  run('git', ['-C', dir, 'reset', '--hard', 'origin/HEAD']);
+}
+
+const url = process.env.RUVNET_CANONICAL_URL || 'https://raw.githubusercontent.com/stuinfla/ruvnet-brain/main/kb';
+// Depth config (--full / --keep) comes from the SHARED per-repo map, overridable per-invocation.
+// Before this, ingest-repo.mjs never passed --full: any repo rebuilt through it silently lost its
+// full-body source indexing (the 2026-07-10 depth-restore run zeroed ruvector this way).
+const FULL = arg('--full', FULL_HINTS[kb] || '');
+const KEEP = arg('--keep', KEEP_DIRS[kb] || '');
+const buildArgs = ['forge-refresh.mjs', '--repo', dir, '--out', '.', '--name', kb, '--canonical-url', url];
+if (FULL) buildArgs.push('--full', FULL);
+if (KEEP) buildArgs.push('--keep', KEEP);
+console.log(`[refresh bge-768] ${kb}${FULL ? ' (--full: ' + FULL.split(',').length + ' prefixes)' : ''}${KEEP ? ' (--keep: ' + KEEP + ')' : ''}`);
+run('node', buildArgs, { cwd: KB, env });
+console.log(`[symbols] ${kb}`);
+try { run('node', ['scripts/build-symbols.mjs', '--name', kb], { cwd: ROOT, env }); } catch { console.log('  (symbols skipped — sparse repo)'); }
+
+const ok = fs.existsSync(path.join(KB, `${kb}.big.rvf`))
+  && fs.existsSync(path.join(KB, `${kb}.passages.jsonl`))
+  && fs.existsSync(path.join(KB, `${kb}.meta.json`));
+console.log(ok
+  ? `\n[done] ${NAME} ingested → searchable now via search_ruvnet (no restart). For capability-confidence, build its primer next.`
+  : `\n[FAIL] ${NAME}: expected stores missing after build.`);
+process.exit(ok ? 0 : 1);

package/scripts/install-npx-witness.sh

@@ -0,0 +1,49 @@
+#!/bin/bash
+# install-npx-witness.sh — install the npx witness as a launchd WatchPaths job. Idempotent.
+#
+# The job fires the instant ~/.npm/_npx changes (any creation/deletion, by anything), runs
+# scripts/npx-witness.sh, and is wrapped in job-heartbeat.sh so it PROVES it ran (standing order:
+# launchd reports exit 0 for a job that never ran — silence is not health).
+#
+# Zero sudo. Zero interference: the job only appends to ~/.claude/logs/npx-witness.log.
+# Uninstall: launchctl bootout gui/$(id -u)/com.ruvnet.npx-witness && rm the plist.
+
+set -euo pipefail
+
+REPO="$(cd "$(dirname "$0")/.." && pwd)"
+LABEL="com.ruvnet.npx-witness"
+PLIST="$HOME/Library/LaunchAgents/$LABEL.plist"
+HB="$REPO/scripts/job-heartbeat.sh"
+WITNESS="$REPO/scripts/npx-witness.sh"
+
+[ -f "$HB" ] || { echo "FAIL: $HB missing"; exit 1; }
+[ -f "$WITNESS" ] || { echo "FAIL: $WITNESS missing"; exit 1; }
+chmod +x "$WITNESS"
+
+cat > "$PLIST" <<EOF
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0"><dict>
+  <key>Label</key><string>$LABEL</string>
+  <key>ProgramArguments</key><array>
+    <string>/bin/sh</string>
+    <string>$HB</string>
+    <string>$LABEL</string>
+    <string>--</string>
+    <string>/bin/bash</string>
+    <string>$WITNESS</string>
+  </array>
+  <key>WatchPaths</key><array><string>$HOME/.npm/_npx</string></array>
+  <key>RunAtLoad</key><false/>
+  <key>StandardOutPath</key><string>$HOME/Library/Logs/npx-witness.out</string>
+  <key>StandardErrorPath</key><string>$HOME/Library/Logs/npx-witness.err</string>
+</dict></plist>
+EOF
+
+launchctl bootout "gui/$(id -u)/$LABEL" 2>/dev/null || true
+launchctl bootstrap "gui/$(id -u)" "$PLIST"
+
+echo "installed: $PLIST"
+echo "loaded:    $(launchctl list | grep -c "$LABEL") (1 = live)"
+echo "log:       ~/.claude/logs/npx-witness.log"
+echo "heartbeat: ~/.cache/ruvnet-brain/heartbeats/$LABEL.json"