ms-ai-architect/scripts/kb-update/check-cosmo-gate.mjs

#!/usr/bin/env node
// check-cosmo-gate.mjs — the RATIFIED R13 persona gate (operator 2026-09-12, alt. A).
//
// Replaces the roadmap's `grep -rl "Cosmo" skills/*/references | wc -l -> 0`, which was
// measured false twice over (it would delete 451 Azure Cosmos DB occurrences, and it is
// unreachable under a heading-only scope because 132 persona occurrences live in prose,
// tables and dialogue). Full derivation: lib/cosmo-persona.mjs header.
//
// THREE CLAUSES
//   G1  persona occurrences on markdown heading lines in the reference corpus = 0
//   G2  every fragment link resolves to a real heading slug, except the anchors on the
//       committed exemption list (pre-existing, not caused by R13)
//   G3  product occurrences unchanged from the committed baseline, and each product-only
//       file still carries zero persona
//
// USAGE
//   node scripts/kb-update/check-cosmo-gate.mjs                  # check (exit 1 on fail)
//   node scripts/kb-update/check-cosmo-gate.mjs --emit-baseline  # write the baseline JSON
//   node scripts/kb-update/check-cosmo-gate.mjs --validate-net   # prove the net both ways
//
// --validate-net is not decoration. A gate whose null result has never been forced to
// the other answer is not a measurement (memory decision-gate-must-be-validated-both-
// ways): it injects a known-positive persona heading into a copy of a real corpus file
// and asserts G1 FAILS, then asserts the known-negative product-only files PASS.
import { readFileSync, writeFileSync, existsSync } from 'node:fs';
import { join, dirname, relative } from 'node:path';
import { fileURLToPath } from 'node:url';
import { globSync } from 'node:fs';
import { classifyCosmo, findDeadAnchors } from './lib/cosmo-persona.mjs';

const __dirname = dirname(fileURLToPath(import.meta.url));
const PLUGIN_ROOT = join(__dirname, '..', '..');
const BASELINE_PATH = join(__dirname, 'data', 'cosmo-gate-baseline.json');

/** Every reference file in the corpus, as repo-relative paths, sorted. */
function refFiles() {
  return globSync('skills/*/references/**/*.md', { cwd: PLUGIN_ROOT }).sort();
}

/** Measure the whole corpus: persona buckets, product total, per-file detail. */
function measure() {
  const files = refFiles();
  const totals = { heading: 0, toc: 0, prose: 0, product: 0 };
  const headingFiles = new Set();
  const perFile = {};
  const deadAnchors = [];
  for (const rel of files) {
    const src = readFileSync(join(PLUGIN_ROOT, rel), 'utf8');
    const c = classifyCosmo(src);
    totals.heading += c.persona.heading;
    totals.toc += c.persona.toc;
    totals.prose += c.persona.prose;
    totals.product += c.product;
    if (c.persona.heading > 0) headingFiles.add(rel);
    perFile[rel] = { persona: c.personaTotal, product: c.product };
    for (const d of findDeadAnchors(src)) deadAnchors.push(`${rel}#${d.anchor}`);
  }
  return { files, totals, headingFiles: [...headingFiles].sort(), perFile, deadAnchors };
}

/** Files carrying Azure Cosmos DB but no persona at all — must stay untouched. */
function productOnlyFiles(m) {
  return m.files.filter((f) => m.perFile[f].product > 0 && m.perFile[f].persona === 0);
}

function emitBaseline() {
  const m = measure();
  const pof = productOnlyFiles(m);
  const baseline = {
    _comment: 'Ratified R13 persona-gate baseline. Measured before the heading transform; '
      + 'G1 drives personaHeadings to 0, G2 and G3 must not move. Regenerate ONLY with a '
      + 'ratified scope change — a silently moved baseline is a gate that proves nothing.',
    measured: new Date().toISOString().slice(0, 10),
    refFiles: m.files.length,
    personaHeadings: m.totals.heading,
    personaHeadingFiles: m.headingFiles.length,
    personaTocOccurrences: m.totals.toc,
    personaProseOccurrences: m.totals.prose,
    productOccurrences: m.totals.product,
    productOnlyFiles: pof,
    productOnlyFileCounts: Object.fromEntries(pof.map((f) => [f, m.perFile[f].product])),
    deadAnchorsExempt: m.deadAnchors,
  };
  writeFileSync(BASELINE_PATH, `${JSON.stringify(baseline, null, 2)}\n`, 'utf8');
  console.log(`baseline skrevet: ${relative(PLUGIN_ROOT, BASELINE_PATH)}`);
  console.log(JSON.stringify(baseline, null, 2));
}

function check() {
  if (!existsSync(BASELINE_PATH)) {
    console.error('FEIL: ingen baseline. Kjoer --emit-baseline foer transformen.');
    process.exit(2);
  }
  const b = JSON.parse(readFileSync(BASELINE_PATH, 'utf8'));
  const m = measure();
  const fails = [];

  // --- G1: no persona left on any heading line ---
  if (m.totals.heading !== 0) {
    fails.push(`G1: ${m.totals.heading} persona-forekomster staar fortsatt paa heading-linjer `
      + `i ${m.headingFiles.length} filer (baseline var ${b.personaHeadings})`);
  }

  // --- G2: no dead fragment links beyond the committed exemptions ---
  const exempt = new Set(b.deadAnchorsExempt);
  const newDead = m.deadAnchors.filter((d) => !exempt.has(d));
  if (newDead.length) {
    fails.push(`G2: ${newDead.length} NYE doede fragmentlenker: ${newDead.slice(0, 10).join(', ')}`);
  }
  const healed = b.deadAnchorsExempt.filter((d) => !m.deadAnchors.includes(d));

  // --- G3: product untouched ---
  if (m.totals.product !== b.productOccurrences) {
    fails.push(`G3: produkt-forekomster ${m.totals.product} != baseline ${b.productOccurrences} `
      + `(differanse ${m.totals.product - b.productOccurrences})`);
  }
  for (const [f, n] of Object.entries(b.productOnlyFileCounts)) {
    const now = m.perFile[f];
    if (!now) { fails.push(`G3: kun-produkt-fila ${f} finnes ikke lenger`); continue; }
    if (now.product !== n) fails.push(`G3: ${f} produkt ${now.product} != baseline ${n}`);
    if (now.persona !== 0) fails.push(`G3: ${f} har faatt ${now.persona} persona-forekomst(er)`);
  }

  console.log('=== R13 persona-gate (ratifisert 2026-09-12, alt. A) ===');
  console.log(`ref-filer                        : ${m.files.length} (baseline ${b.refFiles})`);
  console.log(`G1 persona paa heading-linjer     : ${m.totals.heading}  (baseline ${b.personaHeadings})`);
  console.log(`   persona i TOC-linjer           : ${m.totals.toc}  (baseline ${b.personaTocOccurrences})`);
  console.log(`   persona i proesa/tabell        : ${m.totals.prose}  (baseline ${b.personaProseOccurrences}) `
    + '— UTENFOR R13-scope, bokfoert til R13b/R14');
  console.log(`G2 doede fragmentlenker           : ${m.deadAnchors.length} `
    + `(${exempt.size} unntatt, ${newDead.length} nye, ${healed.length} forsvunnet)`);
  console.log(`G3 produkt-forekomster            : ${m.totals.product}  (baseline ${b.productOccurrences})`);
  console.log(`   kun-produkt-filer urørt        : ${Object.keys(b.productOnlyFileCounts).length}`);
  if (fails.length) {
    console.log('\nGATE FAILED');
    for (const f of fails) console.log(`  x ${f}`);
    process.exit(1);
  }
  console.log('\nGATE PASSED');
}

/** Force the gate to the other answer — a null result nobody could falsify is not one. */
function validateNet() {
  const m = measure();
  let bad = 0;

  // Known-positive: inject a persona heading into a real corpus file's content and
  // assert the classifier sees it. If this cannot fail the gate, the gate is blind.
  const victim = m.files[0];
  const injected = `${readFileSync(join(PLUGIN_ROOT, victim), 'utf8')}\n## For Cosmo\n`;
  const ic = classifyCosmo(injected);
  const base = classifyCosmo(readFileSync(join(PLUGIN_ROOT, victim), 'utf8'));
  if (ic.persona.heading !== base.persona.heading + 1) {
    console.log(`x KJENT-POSITIV: injisert "## For Cosmo" ble IKKE sett `
      + `(${base.persona.heading} -> ${ic.persona.heading})`); bad++;
  } else {
    console.log(`ok KJENT-POSITIV: injisert "## For Cosmo" felte G1 `
      + `(${base.persona.heading} -> ${ic.persona.heading}) i ${victim}`);
  }

  // Known-positive #2: the Norwegian genitive must also be caught, since it is the
  // case the letter-s heuristic gets wrong in BOTH directions.
  const gen = classifyCosmo('### Cosmos tonalitet\n');
  if (gen.persona.heading !== 1) { console.log('x KJENT-POSITIV: genitiv "### Cosmos tonalitet" ikke sett'); bad++; }
  else console.log('ok KJENT-POSITIV: genitiv "### Cosmos tonalitet" felte G1');

  // Known-negative: a product heading must NOT trip the gate.
  const prod = classifyCosmo('### Azure Cosmos DB for GraphRAG\n');
  if (prod.personaTotal !== 0 || prod.product !== 1) {
    console.log(`x KJENT-NEGATIV: produkt-heading feilklassifisert `
      + `(persona ${prod.personaTotal}, produkt ${prod.product})`); bad++;
  } else console.log('ok KJENT-NEGATIV: "### Azure Cosmos DB for GraphRAG" gir 0 persona, 1 produkt');

  // Known-negative #2: the real product-only files must be clean of persona.
  for (const f of productOnlyFiles(m)) {
    const c = classifyCosmo(readFileSync(join(PLUGIN_ROOT, f), 'utf8'));
    if (c.personaTotal !== 0) { console.log(`x KJENT-NEGATIV: ${f} klassifisert med persona`); bad++; }
    else console.log(`ok KJENT-NEGATIV: ${f} — 0 persona, ${c.product} produkt`);
  }

  console.log(bad === 0 ? '\nNETTET VALIDERT BEGGE VEIER' : `\nNETT-VALIDERING FEILET (${bad})`);
  process.exit(bad === 0 ? 0 : 1);
}

const arg = process.argv[2];
if (arg === '--emit-baseline') emitBaseline();
else if (arg === '--validate-net') validateNet();
else check();