Files
CyberGril/web/tests/hel895_real.mjs
T
zonggongandmultica-agent c121c42e39 HEL-896 release completion: the A/B/C budget becomes an executable two-instance split, the real-browser false positive is fixed precisely (product untouched), the SDK gap list is corrected to the honest 41-distribution closure with the pip-download erratum, and release archiving becomes a consistency snapshot - (1) budget split: the manager's independent review reproduced on a temp store that writing A's total_requests=55 into the shared product ledger blocks B's mem0_extract and C's dialogue_main with budget_total_requests_exhausted:55/55 while budget_total/request_usage/token_usage have no stage or authorization-source column; docs/budget-package-hel895.md gains §6 making the split executable: A lives ONLY on the isolated 8360 team-test instance with its own DB and models.yaml (55/500000 one-shot totals plus same-value daily category caps), the user instance NEVER carries one-shot totals (its budget_total rows are record-only, no gate), A is sealed by stop+consistent-snapshot archiving with budget_total never zeroed and restart/archive proving the seal holds, old test data has no import path into the user instance (release checklist forbids copying the 8360 data dir), B starts only after baiqizhi approval writes llm_cap/embed_cap per the 500+180 formula with the authority frozen prefix, the real-Mem0 smoke is counted INSIDE B's 60/180, B cap-hit queues batches (no retry burn, no lost rows, FTS fallback, chat uninterrupted) and first-week end requires a fresh approval to continue, and §6.4 gives one responsible limited recommendation (A 55/500k isolated, B 60 LLM/180 embeds, C daily table) plus errata for the two unsafe original claims ('A shares C's daily account' and 'WAL single file copies whole'); tools/verify_budget_split_hel896.py 8/8 on temp stores and stand-ins only (S0 exact-shape repro of the manager's finding, S1 team instance consumes the §1 step plan to 55 then hard-stops with the tightest gate winning - life daily 8/8 rejects before the total, S2 seal survives restart and backup-API archive, S3 user instance B/C unblocked after A with dialogue 60/semantic 40/daily-token stacking and rejections booking nothing, S4 the misconfigured-total guard on the user instance, S5 repo configs and authoritative facts untouched, real caps all stay 0); (2) browser false positive fixed without touching any product file: web/tests/hel895_real.mjs B3 used '!!stage || any visible button' and clicked the FIRST button containing 回 which is the 日子与回忆 home nav button, so stage=false still PASSed and the screenshot stayed on days - the verbatim original rerun reproduces the false 7/7 with B3 stage=false (docs/rework/hel896/browser-before/), and the revised test passes 9 checks by precise interaction: B2 asserts the letter card carries her exact words plus the talk-nav unread aria-label, B3a waits for the VISIBLE reach card (aside.reach.show, aria-hidden=false, non-zero box, exact words) over days, B3b clicks only THAT card's 回她 button and strictly asserts the conversation stage .conversation .sr contains the reach-out text with the unread mark cleared, B4/B5 prove dedup by server message id (data-mid in the record panel: zero duplicates, gap and live reach-outs each exactly once), B5 asserts /api/health epoch bumps exactly +1 (1->2) before rechecking, B6 counts the page's OWN permissions.query/getUserMedia calls from first document load (0/0, instrumented via Page.addScriptToEvaluateOnNewDocument after Page.enable), B7 runs the service under a temp loopback models config (CYBERGRIL_MODELS, all caps 0) plus a sitecustomize socket guard that REJECTS and eagerly reports non-loopback connects (0 attempts, 0 non-loopback DNS lookups, report survives signal death) - under precise interaction the stage really plays, unread really clears, dedup and epoch semantics really hold: it was a test defect, not a product defect, so no interface change and nothing to escalate; (3) SDK honesty: tools/audit_sdk_offline_hel896.py (read-only, zero downloads/installs/inference) computes the 3.11-Linux closure from the installed 3.10 pinned metadata - 55 dists, 41 missing from the product venv (28 pure + 13 native incl. grpcio/numpy/onnxruntime/tokenizers/sqlalchemy/protobuf), all three top-level pins are py3-none-any (no 'cp311 wheel' needed for them), mem0ai's installed direct_url is the local fixed source hel857-src and hel857-data/mem0-2c81af3.tar.gz hashes exactly to the lock's 9dcef930...dda23 (an index mem0ai==2.2.1 wheel proves nothing), the wheel built during hel857 lived only in an ephemeral cache and hatchling is not offline-available, exactly 3 of the 41 are cache-reusable today for 3.11 (greenlet 3.5.6 cp311, pillow 12.3.0 cp311, truststore 0.10.5 py3-none-any - version+tag verified), the fastembed model cache is measured (Qdrant/bge-small-zh-v1.5 + bm25, ~92MB, placement = instance data dir copy), and the seq48-51 pip event is settled by cache forensics: uv's pip-26.2.2 wheel entry (files.pythonhosted.org, sha256 fd7f4dfa...) was CREATED 2026-10-27 02:10:04 inside the HEL-895 run window as the ONLY new cache entry, so that step DID fetch from the network - the prior card's zero-download claim is corrected to 'download' by the conservative reading (byte-level download vs conditional request is indistinguishable from cache evidence) while the lock install itself added no new cache entries; deploy/hel895-release-candidate.md §4 is rewritten to this accurate list, all future installs forced offline/no-index; (4) release archiving: tools/release_snapshot_hel896.py implements the two safe modes (stop+wal_checkpoint(TRUNCATE)+copy, or online SQLite backup API), writes per-file sha256 manifests carrying the exact code commit, per-table row-count verification and a verify-mode rollback drill; its selftest reproduces the manager's WAL finding (online naive copy of the main .db: 'no such table: companion') and proves both fixed modes complete (backup API and checkpoint copies both hold companion=1/chat_message=1, archive hashes+rows verify, commit 04eff61 recorded); §5 of the release doc now carries the copy-paste commands covering the product DB, semantic vector root, history root and the authoritative embed ledger, archives are never deleted, and the section explicitly does NOT claim R6 complete backup; regression on this head: pytest 299/299 (3.11.16 venv + requirements.lock offline), ruff clean (src/tests/tools), web typecheck 0 errors / lint 0 errors 1 pre-existing warning / vitest 55/55 / build passes with CSS still index-DdDYzR5S.css, pinned-SDK reruns all green (close-ownership 16/16, shutdown-queued 12/12, inflight 6/6, semantic-recovery 7/7, mem0-real 12/12, reports in docs/rework/hel896/regression/), check_release_offline R1-R7 passes, real-HTTP pagination 7/7 rerun saved as docs/rework/hel896/http-pagination-rerun.json with the original hel895 evidence byte-untouched, requirement checks 6/6 (origin/main, de8874a3803b, 152479b, each with and without handoff-HEL-896.json); handoff-HEL-896.json written; zero real LLM, zero real embeddings, zero new downloads beyond existing caches, authoritative ledgers and freezes (39/60, 498/500, 2 frozen, HEL-835 gate) untouched, not deployed, PR#11 appended-not-merged, no visual changes, R5/R6/R7 honestly deferred
Co-authored-by: multica-agent <github@multica.ai>
2026-10-27 03:06:23 +08:00

212 lines
18 KiB
JavaScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
// HEL-895 real-server browser check, HEL-896 revision: drives headless Chromium over CDP against the REAL
// product API (python -m cybergril.api serving web/dist), not the mock. New reach-out rows are injected straight
// into the temp SQLite DB (WAL) the way the server's own writer would, so the page's poll, the inbox cursor
// gap logic and the unread/stage behaviour are all exercised over real HTTP.
//
// HEL-896 fixes the false positives the manager's independent review proved (docs/rework/hel896/):
// - B3 clicked "the first button containing 回" (the 日子与回忆 nav button) and passed on `stage || any
// visible button`. Now it navigates to 日子, waits for the VISIBLE reach card (aside.reach.show) carrying
// exactly this reach-out's words, clicks THAT card's 回她 button, and strictly asserts the conversation
// stage text (.conversation .sr) plus the unread mark clearing. No fallback.
// - B4/B5 dedup evidence is by server message id (data-mid in the record panel), not "1–2 full-page text
// matches"; B5 asserts the restart epoch really is +1 (via /api/health) before re-checking the record.
// - B6 counts the page's own Permissions API / getUserMedia CALLS (instrumented before load), not just states.
// - The service runs with a temp loopback-only models config (CYBERGRIL_MODELS, all caps 0) and a
// sitecustomize socket guard that REJECTS and counts any non-loopback connect, so "zero non-loopback
// outbound" is enforced and evidenced (netguard.json), not assumed. The default /api/health provider probe
// therefore hits a closed loopback port and reports llm.ok=false — that is the isolation working.
// Usage: node tests/hel895_real.mjs <chrome> <venvPython> <repoRoot> <outDir> [port]
import { spawn } from 'node:child_process';
import { mkdirSync, writeFileSync } from 'node:fs';
import { join } from 'node:path';
import { tmpdir } from 'node:os';
import { DatabaseSync } from 'node:sqlite';
const [chrome, venv, root, outArg, portArg] = process.argv.slice(2);
const port = Number(portArg || 8398), base = `http://127.0.0.1:${port}`, cdpPort = 9498;
const out = join(root, outArg || 'docs/rework/hel896/browser');
mkdirSync(out, { recursive: true });
const work = join(tmpdir(), `hel895-real-${Date.now()}`);
mkdirSync(work, { recursive: true });
const dbPath = join(work, 'companion.db');
const results = [];
const check = (name, ok, detail = '') => { results.push(ok); console.log(`${ok ? 'PASS' : 'FAIL'} ${name}${detail ? ` — ${detail}` : ''}`); };
const sleep = ms => new Promise(r => setTimeout(r, ms));
// ── HEL-896 isolation: loopback-only provider config (caps all 0) + socket guard that rejects non-loopback. ──
const loopbackBase = 'http://127.0.0.1:9/v1'; // nothing listens there: provider probes fail fast on loopback
writeFileSync(join(work, 'models-loopback.yaml'), [
'# HEL-896 browser-check isolation config: loopback-only provider, every budget cap 0 (fail-closed).',
'tiers:',
' high: {provider: openai_compat, model: loopback-closed, base_url: "' + loopbackBase + '", api_key_env: null, timeout: 2.0, temperature: 0.1}',
' standard: {provider: openai_compat, model: loopback-closed, base_url: "' + loopbackBase + '", api_key_env: null, timeout: 2.0, temperature: 0.1}',
' low: {provider: openai_compat, model: loopback-closed, base_url: "' + loopbackBase + '", api_key_env: null, timeout: 2.0, temperature: 0.1}',
'budget:',
' daily_tokens: 2000000',
' daily_requests: {dialogue: 0, life: 0, background: 0, semantic: 0}',
'embedding: {provider: openai_compat, model: loopback-closed, base_url: "' + loopbackBase + '", api_key_env: null, dim: 1024}',
'asr: {provider: scripted, model: none, language: zh}',
'tts: {provider: silence, model: none}',
'vad: {provider: energy, model: none}',
'turn: {provider: heuristic, silence_min_ms: 150, silence_hard_end_ms: 800}',
'paralinguistic: {provider: scripted, model: none, enabled: false}',
'avatar: {route: none, renderer: none, lipsync: envelope}',
''].join('\n'));
const guard = join(work, 'guard');
mkdirSync(guard, { recursive: true });
// The guard writes its report EAGERLY (import time + every event, tmp+rename): a uvicorn SIGTERM exit can die
// by signal without running atexit, and a missing report must never be read as "zero attempts".
writeFileSync(join(guard, 'sitecustomize.py'), [
'import json, os, socket, tempfile',
"LOOP = {'127.0.0.1', '::1', 'localhost'}",
"state = {'guard_loaded': True, 'nonloopback_connect_attempts': 0, 'nonloopback_dns_lookups': 0, 'rejected': []}",
'def _host(a):\n try:\n return a[0]\n except Exception:\n return str(a)',
'def _dump():\n try:\n d = os.path.dirname(os.environ["NETGUARD_OUT"])\n fd, tmp = tempfile.mkstemp(dir=d)\n with os.fdopen(fd, "w") as f:\n json.dump(state, f)\n os.replace(tmp, os.environ["NETGUARD_OUT"])\n except Exception:\n pass',
'_connect = socket.socket.connect',
'def connect(self, address):\n if _host(address) not in LOOP:\n state["nonloopback_connect_attempts"] += 1\n state["rejected"].append(str(address))\n _dump()\n raise ConnectionRefusedError("non-loopback blocked: " + str(address))\n return _connect(self, address)',
'socket.socket.connect = connect',
'_create = socket.create_connection',
'def create_connection(address, *a, **k):\n if _host(address) not in LOOP:\n state["nonloopback_connect_attempts"] += 1\n state["rejected"].append(str(address))\n _dump()\n raise ConnectionRefusedError("non-loopback blocked: " + str(address))\n return _create(address, *a, **k)',
'socket.create_connection = create_connection',
'_gai = socket.getaddrinfo',
'def getaddrinfo(host, *a, **k):\n if host not in LOOP:\n state["nonloopback_dns_lookups"] += 1\n _dump()\n return _gai(host, *a, **k)',
'socket.getaddrinfo = getaddrinfo',
'_dump()',
''].join('\n'));
// ── Seed: companion + 5 history rows (one old reach-out) — the same shape as the HEL-875 mock history. ──
const T0 = Date.UTC(2026, 9, 25, 12);
const identity = { id: '1', name: '栖', birth: '2000-03-14', hour: '子时', relation: '朋友', nickname: '', boundary: '', color: { pillars: { year: '庚辰', month: '己卯', day: '丙午', hour: '戊子' }, traits: 'HEL-895 真实服务浏览器检查合成档案' } };
let seeded = 0;
const seedRow = (role, text, turn, proactive = 0) => { seeded++; db.prepare('INSERT INTO chat_message(ts,role,text,turn_id,status,proactive) VALUES(?,?,?,?,?,?)').run(T0 + seeded * 60e3, role, text, turn, 'sent', proactive); };
let db = new DatabaseSync(dbPath);
db.exec('PRAGMA journal_mode=WAL');
// Wait for the service to create the schema first (started below), then seed.
const env = { ...process.env, CYBERGRIL_DB: dbPath, PYTHONPATH: guard + ':' + join(root, 'src'), NETGUARD_OUT: join(work, 'netguard.json'), CYBERGRIL_MODELS: join(work, 'models-loopback.yaml'), CYBERGRIL_R2_ENABLED: '0', CYBERGRIL_SEMANTIC: 'off', CYBERGRIL_PORT: String(port), CYBERGRIL_WEB: join(root, 'web', 'dist') };
let svc = spawn(venv, ['-m', 'cybergril.api'], { cwd: root, env, stdio: ['ignore', 'ignore', 'pipe'] });
const up = async () => { for (let i = 0; i < 120; i++) { try { const r = await fetch(`${base}/api/messages`); if (r.status === 404) return true; } catch { await sleep(500); } } return false; };
if (!await up()) { console.error('service did not come up'); process.exit(1); }
db.exec('PRAGMA journal_mode=WAL');
db.prepare('INSERT INTO companion VALUES(1,?)').run(JSON.stringify(identity));
seedRow('user', '下雨天的时候,你会想听什么?', 't-1');
seedRow('companion', '会想把窗开一条缝,听雨落在铁皮檐上。', 't-1');
seedRow('companion', '昨晚的月亮很亮,想起你说过喜欢夜里散步。', 'reach-old', 1);
seedRow('user', '今天加班到九点,好累。', 't-2');
seedRow('companion', '那先别急着做别的,喝口热水。', 't-2');
const rowIdOf = text => db.prepare('SELECT id FROM chat_message WHERE text=? ORDER BY id DESC LIMIT 1').get(text)?.id;
const restart = async () => { svc.kill('SIGTERM'); await new Promise(r => svc.on('exit', r)); svc = spawn(venv, ['-m', 'cybergril.api'], { cwd: root, env, stdio: ['ignore', 'ignore', 'pipe'] }); for (let i = 0; i < 120; i++) { try { const r = await fetch(`${base}/api/health`); if (r.ok) return await r.json(); } catch { await sleep(500); } } return null; };
// ── Chromium over CDP (same driver style as tests/hel875_shots.mjs). ──
const profile = join(tmpdir(), `hel895-chrome-${Date.now()}`);
const browser = spawn(chrome, ['--headless=new', `--remote-debugging-port=${cdpPort}`, `--user-data-dir=${profile}`, '--no-first-run', '--hide-scrollbars', '--force-color-profile=srgb', 'about:blank'], { stdio: 'ignore' });
let targets;
for (let i = 0; i < 60 && !targets; i++) { try { targets = await (await fetch(`http://127.0.0.1:${cdpPort}/json`)).json(); } catch { await sleep(250); } }
const ws = new WebSocket(targets.find(t => t.type === 'page').webSocketDebuggerUrl);
await new Promise(r => ws.addEventListener('open', r, { once: true }));
let seq = 0; const waiting = new Map();
ws.addEventListener('message', ({ data }) => { const m = JSON.parse(data); if (m.id && waiting.has(m.id)) { const { ok, ko } = waiting.get(m.id); waiting.delete(m.id); m.error ? ko(new Error(m.error.message)) : ok(m.result); } });
const cdp = (method, params = {}) => new Promise((ok, ko) => { const id = ++seq; waiting.set(id, { ok, ko }); ws.send(JSON.stringify({ id, method, params })); });
const js = async expr => { const r = await cdp('Runtime.evaluate', { expression: expr, returnByValue: true, awaitPromise: true }); if (r.exceptionDetails) throw new Error(r.exceptionDetails.exception?.description || 'js error'); return r.result.value; };
const shot = async name => { const { data } = await cdp('Page.captureScreenshot', { format: 'png' }); writeFileSync(join(out, `real-${name}.png`), Buffer.from(data, 'base64')); };
const read = async () => { await js('document.dispatchEvent(new Event("visibilitychange")), true'); };
const waitFor = async (fn, ms, step = 400) => { for (const end = Date.now() + ms; Date.now() < end;) { const v = await fn(); if (v) return v; await sleep(step); } return null; };
// HEL-896: count the page's OWN permission API calls from the very first document load.
await cdp('Page.enable');
await cdp('Page.addScriptToEvaluateOnNewDocument', { source: `(() => { let q = 0, g = 0; const qp = navigator.permissions.query.bind(navigator.permissions);
navigator.permissions.query = d => { q++; return qp(d); };
const gm = navigator.mediaDevices && navigator.mediaDevices.getUserMedia && navigator.mediaDevices.getUserMedia.bind(navigator.mediaDevices);
if (gm) navigator.mediaDevices.getUserMedia = d => { g++; return gm(d); };
window.__permCounts = () => ({ query: q, getUserMedia: g }); })()` });
const talkUnread = () => js(`(() => { const b = [...document.querySelectorAll('.nav-item')].find(e => e.textContent.trim() === '交谈');
return b ? { unread: b.classList.contains('unread'), label: b.getAttribute('aria-label') || '' } : null })()`);
const navTo = async label => { await js(`[...document.querySelectorAll('.nav-item')].find(e => e.textContent.trim() === ${JSON.stringify(label)})?.click(), true`); await sleep(1200); };
try {
// B1 — the built web boots against the real service.
await cdp('Page.navigate', { url: `${base}/` }); await sleep(2500);
const ready = await js(`!!document.querySelector('.conversation, .stage, [class*=talk], [class*=days]') && !document.body.innerText.includes('服务尚未接通')`);
check('B1_real_service_boot', ready, await js(`document.body.innerText.slice(0, 40)`));
await shot('boot');
// B2 — a reach-out lands while the page is open on 此刻: the letter carries her exact words, the talk nav marks unread.
seedRow('companion', '窗外的雨停了,想跟你说一声。', `reach-live-${Date.now()}`, 1);
await read();
const letter = await waitFor(async () => { await read(); return js(`(() => { const b = document.querySelector('button.letter:not(.read)');
return b && b.querySelector('p')?.textContent.includes('窗外的雨停了') ? b.querySelector('p').textContent : null })()`); }, 30000, 1500);
const unreadBefore = await waitFor(async () => { const u = await talkUnread(); return u && u.unread ? u : null; }, 8000);
check('B2_reach_card_over_real_http', !!letter && !!unreadBefore, `letter=${!!letter} unread=${unreadBefore?.label || 'none'}`);
await shot('card');
// B3 — the reach card speaks over 日子 and its OWN 回她 button opens the conversation: the reach-out plays on
// the stage and the unread mark clears. Precise: no "any button with 回", no `|| unread` fallback.
await navTo('日子');
const card = await waitFor(() => js(`(() => { const a = document.querySelector('aside.reach.show');
return a && a.getAttribute('aria-hidden') === 'false' && a.getBoundingClientRect().width > 0 && a.querySelector('p')?.textContent.includes('窗外的雨停了') ? a.querySelector('p').textContent : null })()`), 8000);
check('B3a_visible_reach_card', !!card, card ? '' : 'aside.reach.show not visible');
const clicked = await js(`(() => { const a = document.querySelector('aside.reach.show'); if (!a) return false;
const b = [...a.querySelectorAll('.acts button')].find(e => e.textContent.trim() === '回她'); if (!b) return false; b.click(); return true })()`);
await sleep(1500);
const stage = await waitFor(() => js(`(document.querySelector('.conversation .sr')?.textContent || '').includes('窗外的雨停了')`), 8000);
const unreadAfter = await talkUnread();
check('B3b_reply_opens_stage_plays_reachout', !!card && !!clicked && !!stage && !unreadAfter?.unread,
`card=${!!card} clicked=${!!clicked} stage=${!!stage} unreadAfter=${unreadAfter?.unread}`);
await shot('stage');
// B4 — long absence: 260 bulk rows land ABOVE one new reach-out (gap regression, HEL-880): the bounded first
// read (5×50) must not skip her; the next read closes the gap and hands her over — proven by server message id.
seedRow('companion', '刚看到一只猫从窗台跳下去。', `reach-gap-${Date.now()}`, 1);
for (let i = 0; i < 260; i++) seedRow(i % 2 ? 'companion' : 'user', `别处的第${i + 1}句`, `bulk-${Date.now()}-${i}`);
const gapId = rowIdOf('刚看到一只猫从窗台跳下去。'), liveId = rowIdOf('窗外的雨停了,想跟你说一声。');
const badge = await waitFor(async () => { await read(); const u = await talkUnread(); return u && u.unread ? u : null; }, 40000, 1500);
check('B4a_gap_read_hands_reachout', !!badge, `gapId=${gapId} ${badge?.label || 'no badge'}`);
await shot('gap');
// Leaving the conversation merges the queued reach-out into the record (the one unread rule); the record's
// data-mid entries are the server's own ids, so dedup is proven per message id, not by counting text matches.
await navTo('日子'); await navTo('交谈');
await js(`document.querySelector('button[aria-label="完整记录"]')?.click(), true`); await sleep(1200);
const recordIds = () => js(`[...document.querySelectorAll('.record [data-mid]')].map(e => e.getAttribute('data-mid'))`);
let ids = await waitFor(async () => { const v = await recordIds(); return v && v.includes(String(gapId)) ? v : null; }, 10000);
const dupCount = ids ? ids.length - new Set(ids).size : -1;
const gapHits = ids ? ids.filter(i => i === String(gapId)).length : 0;
const liveHits = ids ? ids.filter(i => i === String(liveId)).length : 0;
check('B4b_no_duplicate_by_message_id', !!ids && dupCount === 0 && gapHits === 1 && liveHits === 1,
`rows=${ids?.length || 0} dups=${dupCount} gapId×${gapHits} liveId×${liveHits}`);
await shot('record');
await js(`document.querySelector('button[aria-label="关闭记录"]')?.click(), true`); await sleep(600);
// B5 — server restart: the epoch really bumps by exactly one, and the same rows are not added again.
const epochBefore = (await (await fetch(`${base}/api/health`)).json()).epoch;
const healthAfter = await restart(); await sleep(800);
const epochAfter = healthAfter?.epoch;
await read(); await sleep(2500); await read(); await sleep(1500);
await js(`document.querySelector('button[aria-label="完整记录"]')?.click(), true`); await sleep(1200);
ids = await recordIds();
const dup2 = ids ? ids.length - new Set(ids).size : -1;
const gap2 = ids ? ids.filter(i => i === String(gapId)).length : 0;
check('B5_restart_epoch_plus1_no_dup', epochAfter === epochBefore + 1 && dup2 === 0 && gap2 === 1,
`epoch=${epochBefore}->${epochAfter} rows=${ids?.length || 0} dups=${dup2} gapId×${gap2}`);
// B6 — the page never called the Permissions API or getUserMedia at all (instrumented from first load).
const counts = await js('window.__permCounts()');
const mic = await js(`navigator.permissions.query({name:'microphone'}).then(r => r.state)`);
const cam = await js(`navigator.permissions.query({name:'camera'}).then(r => r.state)`);
check('B6_zero_permission_api_calls', !!counts && counts.query === 0 && counts.getUserMedia === 0,
`counts=${JSON.stringify(counts)} mic=${mic} cam=${cam}`);
} catch (e) {
console.error('driver error:', e.message);
results.push(false);
} finally {
// HEL-896: the service must EXIT so the socket guard's atexit report exists; non-loopback budget is then read
// from its own counter file, not assumed.
try { svc.kill('SIGTERM'); } catch { /* already gone */ }
await new Promise(r => { if (svc.exitCode !== null || svc.signalCode) return r(); svc.on('exit', r); setTimeout(r, 8000); });
let netguard = null;
try { netguard = JSON.parse((await import('node:fs')).readFileSync(join(work, 'netguard.json'), 'utf8')); } catch { /* missing report = honest fail below */ }
check('B7_zero_nonloopback_outbound', !!netguard && netguard.nonloopback_connect_attempts === 0,
netguard ? JSON.stringify(netguard) : 'netguard.json missing');
writeFileSync(join(out, 'real-browser.json'), JSON.stringify({ ok: results.every(Boolean), checks: results.length, fail_count: results.filter(v => !v).length, netguard: netguard || 'missing', service_epoch_isolation: 'CYBERGRIL_MODELS loopback, caps 0, socket guard rejecting non-loopback' }, null, 1));
try { browser.kill(); } catch { /* already gone */ }
ws.close();
}
process.exit(results.every(Boolean) ? 0 : 1);