roadmap.html


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971

<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta name="theme-color" content="#000000">
<title>newOS4 Roadmap — Self-Improving Autonomous OS</title>
<style>
:root {
    --bg-color: #000000;
    --bg-secondary: #0a0a0a;
    --bg-tertiary: #111111;
    --text-color: #00ff00;
    --text-muted: #888888;
    --accent-color: #00aa00;
    --border-color: #005500;
    --error-color: #ff5555;
    --warning-color: #ffaa00;
    --success-color: #00ff00;
    --info-color: #55aaff;
    --phase-color: #00aa00;
    --danger-color: #ff4444;
}
* { box-sizing: border-box; margin: 0; padding: 0; }
body {
    background: var(--bg-color);
    color: var(--text-color);
    font-family: 'Cascadia Code', 'JetBrains Mono', 'Fira Code', 'Consolas', 'Courier New', monospace;
    line-height: 1.6;
    font-size: 14px;
    padding: 0;
    overflow-x: hidden;
}
.container { max-width: 1100px; margin: 0 auto; padding: 20px; }

/* Header */
header {
    background: var(--bg-secondary);
    border-bottom: 2px solid var(--border-color);
    padding: 30px 20px;
    text-align: center;
}
header h1 {
    font-size: 28px;
    color: var(--text-color);
    text-shadow: 0 0 10px rgba(0,255,0,0.3);
    letter-spacing: 1px;
}
header .subtitle { color: var(--text-muted); font-size: 13px; margin-top: 8px; }
header .ver-badge {
    display: inline-block;
    background: var(--bg-tertiary);
    border: 1px solid var(--accent-color);
    border-radius: 4px;
    padding: 3px 12px;
    margin: 5px 3px;
    font-size: 12px;
    color: var(--accent-color);
}
header .ver-badge.target { color: var(--warning-color); border-color: var(--warning-color); }

/* Status bar */
.status-bar {
    display: flex;
    gap: 10px;
    justify-content: center;
    flex-wrap: wrap;
    margin: 15px 0;
}
.status-pill {
    padding: 4px 14px;
    border-radius: 20px;
    font-size: 12px;
    font-weight: bold;
    border: 1px solid;
}
.status-live { background: rgba(0,255,0,0.1); border-color: var(--success-color); color: var(--success-color); }
.status-pending { background: rgba(255,170,0,0.1); border-color: var(--warning-color); color: var(--warning-color); }
.status-blocked { background: rgba(255,85,85,0.1); border-color: var(--error-color); color: var(--error-color); }

/* Sections */
section {
    background: var(--bg-secondary);
    border: 1px solid var(--border-color);
    border-radius: 6px;
    padding: 25px;
    margin: 20px 0;
}
section h2 {
    font-size: 18px;
    color: var(--accent-color);
    margin-bottom: 15px;
    padding-bottom: 8px;
    border-bottom: 1px solid var(--border-color);
}
section h3 {
    font-size: 15px;
    color: var(--text-color);
    margin: 15px 0 8px;
}

/* Tables */
table {
    width: 100%;
    border-collapse: collapse;
    margin: 10px 0;
    font-size: 13px;
}
th {
    background: var(--bg-tertiary);
    color: var(--accent-color);
    text-align: left;
    padding: 8px 12px;
    border: 1px solid var(--border-color);
    font-weight: normal;
    text-transform: uppercase;
    letter-spacing: 0.5px;
    font-size: 11px;
}
td {
    padding: 8px 12px;
    border: 1px solid var(--border-color);
    color: var(--text-color);
    vertical-align: top;
}
tr:nth-child(even) td { background: var(--bg-tertiary); }
td.center, th.center { text-align: center; }
td.muted { color: var(--text-muted); }
td.pass { color: var(--success-color); }
td.fail { color: var(--error-color); }
td.warn { color: var(--warning-color); }
td.info { color: var(--info-color); }

/* Pipeline diagram */
.pipeline-box {
    background: var(--bg-color);
    border: 1px solid var(--border-color);
    border-radius: 6px;
    padding: 20px;
    overflow-x: auto;
    font-size: 12px;
    margin: 10px 0;
}
.pipeline-box pre {
    color: var(--text-color);
    white-space: pre;
}
.pipeline-box .step { color: var(--accent-color); }
.pipeline-box .new { color: var(--warning-color); font-weight: bold; }
.pipeline-box .gate { color: var(--danger-color); font-weight: bold; }
.pipeline-box .ok { color: var(--success-color); }

/* Phase cards */
.phase-card {
    background: var(--bg-tertiary);
    border: 1px solid var(--border-color);
    border-radius: 6px;
    padding: 18px;
    margin: 12px 0;
}
.phase-card:hover { border-color: var(--accent-color); }
.phase-header {
    display: flex;
    justify-content: space-between;
    align-items: center;
    margin-bottom: 10px;
}
.phase-num {
    font-size: 11px;
    color: var(--accent-color);
    font-weight: bold;
    letter-spacing: 1px;
}
.phase-ver {
    font-size: 11px;
    color: var(--warning-color);
    border: 1px solid var(--warning-color);
    border-radius: 3px;
    padding: 2px 8px;
}
.phase-effort {
    font-size: 11px;
    color: var(--text-muted);
}
.phase-title { font-size: 14px; color: var(--text-color); margin-bottom: 8px; }

/* Checklist */
.checklist { list-style: none; padding: 0; }
.checklist li {
    padding: 5px 0 5px 28px;
    position: relative;
    font-size: 13px;
}
.checklist li::before {
    content: '';
    position: absolute;
    left: 0;
    top: 5px;
    width: 18px;
    height: 18px;
    border: 1px solid var(--border-color);
    border-radius: 3px;
    background: var(--bg-color);
}
.checklist li.done::before {
    background: var(--success-color);
    border-color: var(--success-color);
}
.checklist li.done::after {
    content: '✓';
    position: absolute;
    left: 4px;
    top: 3px;
    color: var(--bg-color);
    font-weight: bold;
    font-size: 13px;
}
.checklist li.pending::before {
    border-color: var(--warning-color);
}
.checklist li.pending::after {
    content: '◯';
    position: absolute;
    left: 3px;
    top: 2px;
    color: var(--warning-color);
    font-size: 14px;
}

/* Confidence gate */
.confidence-gate {
    display: flex;
    gap: 0;
    margin: 15px 0;
    border-radius: 6px;
    overflow: hidden;
    border: 1px solid var(--border-color);
}
.confidence-block {
    flex: 1;
    padding: 12px;
    text-align: center;
    font-size: 12px;
}
.confidence-block .score { font-size: 20px; font-weight: bold; display: block; }
.confidence-block .label { font-size: 11px; text-transform: uppercase; letter-spacing: 0.5px; }
.cb-auto { background: rgba(0,255,0,0.08); color: var(--success-color); border-right: 1px solid var(--border-color); }
.cb-auto-log { background: rgba(0,170,0,0.08); color: var(--accent-color); border-right: 1px solid var(--border-color); }
.cb-human { background: rgba(255,170,0,0.08); color: var(--warning-color); border-right: 1px solid var(--border-color); }
.cb-abort { background: rgba(255,85,85,0.08); color: var(--error-color); }

/* Gap analysis */
.gap-card {
    background: var(--bg-tertiary);
    border-left: 3px solid var(--border-color);
    padding: 14px;
    margin: 10px 0;
    border-radius: 0 4px 4px 0;
}
.gap-card.p0 { border-left-color: var(--error-color); }
.gap-card.p1 { border-left-color: var(--warning-color); }
.gap-card.p2 { border-left-color: var(--info-color); }
.gap-id { font-size: 11px; color: var(--text-muted); font-weight: bold; }
.gap-title { font-size: 14px; color: var(--text-color); margin: 3px 0 5px; }
.gap-desc { font-size: 12px; color: var(--text-muted); margin-bottom: 5px; }
.gap-fix {
    font-size: 12px;
    color: var(--accent-color);
    background: var(--bg-color);
    border: 1px solid var(--border-color);
    border-radius: 3px;
    padding: 8px 10px;
    margin-top: 5px;
    white-space: pre-wrap;
}

/* Test scenarios */
.test-scenario {
    background: var(--bg-tertiary);
    border: 1px solid var(--border-color);
    border-radius: 4px;
    padding: 12px;
    margin: 8px 0;
    font-size: 12px;
}
.test-id { font-weight: bold; color: var(--warning-color); }
.test-desc { color: var(--text-color); margin: 3px 0; }
.test-expected { color: var(--text-muted); }

/* Open questions */
.open-q {
    background: var(--bg-tertiary);
    border: 1px solid var(--border-color);
    border-radius: 4px;
    padding: 14px;
    margin: 10px 0;
}
.open-q .q-num { font-weight: bold; color: var(--warning-color); }
.open-q .q-text { color: var(--text-color); margin-top: 3px; }

/* Footer */
footer {
    text-align: center;
    padding: 30px 20px;
    color: var(--text-muted);
    font-size: 11px;
    border-top: 1px solid var(--border-color);
    margin-top: 30px;
}
footer a { color: var(--info-color); text-decoration: none; }

/* Risk matrix */
.risk-matrix td { font-size: 12px; }
.risk-high { color: var(--error-color); font-weight: bold; }
.risk-med { color: var(--warning-color); }
.risk-low { color: var(--success-color); }

/* Scrollbar */
::-webkit-scrollbar { width: 8px; height: 8px; }
::-webkit-scrollbar-track { background: var(--bg-secondary); }
::-webkit-scrollbar-thumb { background: var(--border-color); border-radius: 4px; }
::-webkit-scrollbar-thumb:hover { background: var(--accent-color); }

/* Responsive */
@media (max-width: 600px) {
    .container { padding: 10px; }
    section { padding: 15px; }
    header h1 { font-size: 20px; }
    table { font-size: 11px; }
    th, td { padding: 5px 6px; }
    .confidence-gate { flex-direction: column; }
    .confidence-block { border-right: none; border-bottom: 1px solid var(--border-color); }
}
</style>
</head>
<body>

<header>
    <h1>newOS4 Roadmap</h1>
    <div class="subtitle">Self-Improving Autonomous OS — Pipeline Review & Implementation Plan</div>
    <div style="margin-top: 10px;">
        <span class="ver-badge">v4.6.03 LIVE</span>
        <span class="ver-badge" style="color: var(--success-color); border-color: var(--success-color);">v4.6.09 SHIPPED</span>
        <span class="ver-badge" style="color: var(--warning-color); border-color: var(--warning-color);">v4.6.10 PLANNED</span>
    </div>
    <div class="status-bar">
        <span class="status-pill status-live">9/9 Tools Live</span>
        <span class="status-pill status-live">7/7 Phases (6 shipped, 1 planned)</span>
        <span class="status-pill status-live">Auto-Commit Active</span>
    </div>
</header>

<div class="container">

<!-- ============ EXECUTIVE SUMMARY ============ -->
<section>
    <h2>📋 Executive Summary</h2>
    <p style="color: var(--text-color); margin-bottom: 12px;">
        The self-improvement pipeline is <strong style="color:var(--success-color)">live at v4.6.09</strong> with a 7-step loop:
        <code style="color:var(--accent-color)">edit → test → health → regression → score → auto-commit (≥70) → verify → learn</code>.
        All 6 original phases shipped. <strong style="color:var(--warning-color)">Phase 7 (v4.6.10)</strong> adds a
        <strong style="color:var(--warning-color)">DebugSupervisorModule</strong> — persistent error database, clustering,
        root-cause analysis, and cross-session auto-fix learning.
    </div>
</section>

<!-- ============ CURRENT STATE ============ -->
<section>
    <h2>✅ Current State — What's Live (v4.6.09)</h2>

    <h3>Pipeline Architecture (verified in browser)</h3>
    <div class="pipeline-box">
<pre><span class="step">ship_module</span>({name, code, tests, commit:true})
  │
  ├─ 1. WRITE    → <span class="step">_mods/Module.js</span> + meta.json
  ├─ 2. TEST     → <span class="step">test_module</span>(name, tests)
  │     ├─ Load from _mods/ via loadModuleFromCode()
  │     ├─ <span class="new">JS error capture ← LIVE</span>
  │     ├─ Check: module loaded OK? → <span class="ok">rollback if not</span>
  │     ├─ Check: commands lost? → <span class="ok">rollback if yes</span>
  │     ├─ Run test assertions [{command, expect}] → <span class="ok">rollback on fail</span>
  │     └─ Smoke test: run first command → <span class="ok">rollback on fail</span>
  ├─ 3. HEALTH   → <span class="step">health_check</span>()
  │     ├─ Core commands: ls, pwd, echo, help, grep, find, tree, cat
  │     ├─ <span class="new">JS error capture ← LIVE</span>
  │     ├─ Module count
  │     ├─ World model version match
  │     └─ Agent status (total/healthy/failed)
  ├─ 4. REGRESS  → <span class="step">_runRegressionSuite</span>(moduleName)
  │     └─ <span class="new">30+ tests across all modules ← LIVE</span>
  ├─ 5. SCORE    → <span class="step">_calculateConfidence()</span>
  │     └─ score ≥ 70 → <span class="ok">auto-approve</span>
  │     └─ score &lt; 70 → human approval required
  ├─ 6. COMMIT   → <span class="step">_executeCommit()</span>
  │     ├─ WebDAV push standalone .js
  │     ├─ Bundle sync (find class block, replace)
  │     ├─ Version bump (main.html + module-loader.js)
  │     ├─ <span class="new">Post-commit syntax check ← LIVE</span>
  │     └─ <span class="new">Auto-revert on syntax error ← LIVE</span>
  ├─ 7. VERIFY   → <span class="new">Post-commit fetch + validate ← LIVE</span>
  └─ 8. LEARN    → <span class="step">_postRunReview</span> + skill creation
        └─ <span class="new">Auto-fix suggestions ← LIVE</span></pre>
    </div>

    <h3>Tools Verified Live</h3>
    <table>
        <thead>
            <tr><th>Tool</th><th class="center">Status</th><th>Notes</th></tr>
        </thead>
        <tbody>
            <tr><td><code>edit_module</code></td><td class="center pass">✅ Live</td><td>Writes to <code>_mods/</code>, auto-loads</td></tr>
            <tr><td><code>test_module</code></td><td class="center pass">✅ Live</td><td>Has assertions + inline rollback</td></tr>
            <tr><td><code>reload_module</code></td><td class="center pass">✅ Live</td><td>Reverts <code>_mods/</code>, reloads from URL</td></tr>
            <tr><td><code>list_modified_modules</code></td><td class="center pass">✅ Live</td><td>Lists pending edits</td></tr>
            <tr><td><code>health_check</code></td><td class="center pass">✅ Live</td><td>Core cmds + world model + agents</td></tr>
            <tr><td><code>diff_module</code></td><td class="center pass">✅ Live</td><td>Line-by-line diff, orig vs new</td></tr>
            <tr><td><code>ship_module</code></td><td class="center pass">✅ Live</td><td>Full pipeline orchestrator with auto-commit (score ≥ 70)</td></tr>
            <tr><td><code>commit_module</code></td><td class="center pass">✅ Live</td><td>Auto-executes if confidence ≥ 70, else creates pending commit</td></tr>
            <tr><td><code>run_command</code></td><td class="center pass">✅ Live</td><td>Captures output via print monkey-patch</td></tr>
        </tbody>
    </table>

    <h3>Terminal Commands (human-facing)</h3>
    <table>
        <thead>
            <tr><th>Command</th><th class="center">Status</th><th>Notes</th></tr>
        </thead>
        <tbody>
            <tr><td><code>commit-approve</code></td><td class="center pass">✅ Live</td><td>Executes pending commit (WebDAV + bundle + version)</td></tr>
            <tr><td><code>commit-deny</code></td><td class="center pass">✅ Live</td><td>Cancels pending commit</td></tr>
            <tr><td><code>commit-status</code></td><td class="center pass">✅ Live</td><td>Shows pending commit details</td></tr>
        </tbody>
    </table>
</section>

<!-- ============ GAP ANALYSIS ============ -->
<section>
    <h2>🔍 Gap Analysis — What's Missing for Full Autonomy</h2>

    <div class="gap-card p0">
        <div class="gap-id">GAP 1 — P0 SAFETY CRITICAL</div>
        <div class="gap-title">No JS Error Capture — <code>_captureConsoleErrors()</code> never coded</div>
        <div class="gap-desc">
            The <code>health_check</code> handler sets <code>results.healthy = allCoreOk && results.versionMatch</code> —
            it does <strong>NOT</strong> check for JavaScript errors. A module edit that introduces a silent
            <code>TypeError</code> in a method that isn't called by the 4 core commands will pass health_check
            and proceed to commit.
        </div>
        <div class="gap-desc"><strong>Impact:</strong> AI can commit a module that throws errors on <code>grep</code>, <code>find</code>, <code>tree</code> without detection.</div>
        <div class="gap-fix">Fix: Implement <code>_captureConsoleErrors(durationMs)</code> using <code>console.error</code> + <code>window.onerror</code> interception. Wire into both <code>test_module</code> (after load) and <code>health_check</code> (add to <code>results.healthy</code> check).</div>
    </div>

    <div class="gap-card p0">
        <div class="gap-id">GAP 2 — P0 COVERAGE CRITICAL</div>
        <div class="gap-title">Regression Suite Too Thin — 15/19 modules have zero tests</div>
        <table style="margin: 8px 0;">
            <tr><th>Module</th><th class="center">Tests</th><th>Module</th><th class="center">Tests</th></tr>
            <tr><td>ShellModule</td><td class="center pass">4</td><td>WebDAVModule</td><td class="center fail">0</td></tr>
            <tr><td>SystemModule</td><td class="center pass">2</td><td>CronModule</td><td class="center fail">0</td></tr>
            <tr><td>OrchestratorModule</td><td class="center pass">1</td><td>ThemeModule</td><td class="center fail">0</td></tr>
            <tr><td>WebX11AIAssistantModule</td><td class="center pass">1</td><td>MindmapModule</td><td class="center fail">0</td></tr>
            <tr><td colspan="2" class="muted">+ 11 more untested</td><td colspan="2" class="muted">CalcModule, MediaModule, VoiceModule, etc.</td></tr>
        </table>
        <div class="gap-desc"><strong>Impact:</strong> AI edits any of 15 untested modules → regression returns <code>{allPassed: true, total: 0}</code> — a false positive. Pipeline commits without verification.</div>
        <div class="gap-fix">Fix: Add 2+ regression tests per module to <code>_regressionTasks()</code>. Data-only change — the array just needs more entries. Target: 30+ tests across all 19 modules.</div>
    </div>

    <div class="gap-card p1">
        <div class="gap-id">GAP 3 — P1 SAFETY NET</div>
        <div class="gap-title">No Post-Commit Verification</div>
        <div class="gap-desc">
            After <code>commit-approve</code> pushes files via WebDAV, there's no verification that the committed
            code actually works on a fresh page load. If the code has a syntax error that passes
            <code>loadModuleFromCode</code> but fails at HTTP fetch time, the next user refresh breaks the system.
        </div>
        <div class="gap-fix">Fix: After WebDAV push, fetch the file via HTTP, validate with <code>new Function(code)</code> (browser-safe syntax check). If validation fails, auto-revert by re-uploading the original code.</div>
    </div>

    <div class="gap-card p1">
        <div class="gap-id">GAP 4 — P1 DECISION GATE</div>
        <div class="gap-title">No Confidence Score</div>
        <div class="gap-desc">
            The pipeline is binary — all tests pass → commit, any test fails → rollback. A 1-line typo fix in
            <code>CalcModule</code> (no regression tests) is treated the same as a 200-line rewrite of
            <code>ShellModule</code> (4 regression tests). Both get <code>commit: true</code>.
        </div>
        <div class="gap-fix">Fix: <code>_calculateConfidence(modName, testResult, healthResult, regResult, changeCount)</code>. Penalize: large changes, no regression tests, critical modules. Bonus: comprehensive AI-written tests. Auto-approve at ≥ 70.</div>
    </div>

    <div class="gap-card p1">
        <div class="gap-id">GAP 5 — P1 RECOVERY</div>
        <div class="gap-title">No Automatic Rollback on Next-Boot Failure</div>
        <div class="gap-desc">
            If a committed change breaks the system on next page load, there's no automatic recovery.
            The user has to manually identify the broken module and revert it.
        </div>
        <div class="gap-fix">Fix: Add a "boot watchdog" — before commit, save version to <code>last-good-version.json</code>. On boot, if core modules fail to register within timeout, auto-fetch and restore last-good versions.</div>
    </div>

    <div class="gap-card p2">
        <div class="gap-id">GAP 6 — P2 AUTONOMY</div>
        <div class="gap-title">Self-Improvement Trigger Not Implemented (Phase G)</div>
        <div class="gap-desc">
            The plan's Phase G called for extending <code>_postRunReview</code> to detect repeated tool failures
            and suggest <code>edit_module</code> fixes. This was never coded. The AI doesn't autonomously
            identify bugs to fix — it only uses the pipeline when told to.
        </div>
        <div class="gap-fix">Fix: Build <code>_toolToModuleMap</code> at init. Extend <code>_postRunReview</code> to detect repeated tool failures → push auto-fix suggestions into next AI turn's context.</div>
    </div>
</section>

<!-- ============ CONFIDENCE GATE ============ -->
<section>
    <h2>🚦 The Confidence Gate (Replaces Human Gate)</h2>
    <p style="color: var(--text-muted); margin-bottom: 12px;">
        Instead of a binary human-approve/deny, the system calculates a confidence score based on
        test coverage, change size, and module criticality. High-confidence changes auto-commit;
        low-confidence changes still require human approval.
    </p>
    <div class="confidence-gate">
        <div class="confidence-block cb-auto">
            <span class="score">≥ 85</span>
            <span class="label">Auto-commit<br>immediately</span>
        </div>
        <div class="confidence-block cb-auto-log">
            <span class="score">70–84</span>
            <span class="label">Auto-commit<br>with log</span>
        </div>
        <div class="confidence-block cb-human">
            <span class="score">50–69</span>
            <span class="label">Human approval<br>required</span>
        </div>
        <div class="confidence-block cb-abort">
            <span class="score">&lt; 50</span>
            <span class="label">Auto-abort</span>
        </div>
    </div>

    <h3>Scoring Formula</h3>
    <div class="pipeline-box">
<pre>Starting score: <span class="ok">100</span>

<span style="color:var(--error-color)">Penalties:</span>
  Change > 50 lines:          -20
  Change > 100 lines:         -20 (additional)
  No regression tests:        -30
  Critical module:            -15  (ShellModule, SystemModule, OrchestratorModule, AIAssistant)
  JS errors detected:         -50

<span style="color:var(--success-color)">Bonuses:</span>
  3+ AI-written tests:        +10

<span style="color:var(--warning-color)">Final score = max(0, min(100, score))</span></pre>
    </div>
</section>

<!-- ============ TARGET ARCHITECTURE ============ -->
<section>
    <h2>🏗️ Target Architecture: v4.6.10 (Full Autonomy)</h2>
    <div class="pipeline-box">
<pre><span class="step">ship_module</span>({name, code, tests, commit:true})
  │
  ├─ 1. WRITE    → <span class="step">_mods/Module.js</span>
  ├─ 2. TEST     → <span class="step">test_module</span>(name, tests)
  │     ├─ Load + <span class="new">JS error capture ← NEW</span>
  │     ├─ Command count check
  │     ├─ Test assertions
  │     └─ Smoke test
  ├─ 3. HEALTH   → <span class="step">health_check</span>()
  │     ├─ Core commands (<span class="new">expanded to 8 ← NEW</span>)
  │     ├─ <span class="new">JS error capture ← NEW</span>
  │     ├─ Module count
  │     ├─ World model version
  │     └─ Agent status
  ├─ 4. REGRESS  → <span class="step">_runRegressionSuite</span>(moduleName)
  │     └─ <span class="new">30+ tests across all modules ← EXPANDED</span>
  ├─ 5. SCORE    → <span class="new">_calculateConfidence() ← NEW</span>
  │     └─ score ≥ 70 → auto-approve
  │     └─ score < 70 → human approval required
  ├─ 6. COMMIT   → <span class="step">_executeCommit()</span> (<span class="new">auto if score ≥ 70 ← NEW</span>)
  │     ├─ WebDAV push
  │     ├─ Bundle sync
  │     ├─ Version bump
  │     ├─ <span class="new">Post-commit syntax check ← NEW</span>
  │     └─ <span class="new">Auto-revert on syntax error ← NEW</span>
  ├─ 7. VERIFY   → <span class="new">Post-commit fetch + validate ← NEW</span>
  └─ 8. LEARN    → <span class="step">_postRunReview</span> + skill creation
        └─ <span class="new">Auto-fix suggestions ← NEW</span></pre>
    </div>
</section>

<!-- ============ IMPLEMENTATION PHASES ============ -->
<section>
    <h2>📅 Implementation Phases</h2>

    <div class="phase-card">
        <div class="phase-header">
            <span class="phase-num">PHASE 1</span>
            <span class="phase-ver" style="color: var(--success-color); border-color: var(--success-color);">✅ v4.6.03 SHIPPED</span>
            <span class="phase-effort">~1.5h</span>
        </div>
        <div class="phase-title">Safety Nets — JS Error Capture</div>
        <ul class="checklist">
            <li class="done">Implement <code>_captureConsoleErrors(durationMs)</code> using console.error + window.onerror interception</li>
            <li class="done">Wire JS error check into <code>test_module</code> (after loadModuleFromCode)</li>
            <li class="done">Wire JS error check into <code>health_check</code> (add to <code>results.healthy</code>)</li>
            <li class="done">Expand core commands in health_check: add <code>grep</code>, <code>find</code>, <code>tree</code>, <code>cat</code></li>
            <li class="done">Sync changes to <code>core-bundle.js</code></li>
            <li class="done">Browser verify: inject error, confirm health_check catches it</li>
        </ul>
    </div>

    <div class="phase-card">
        <div class="phase-header">
            <span class="phase-num">PHASE 2</span>
            <span class="phase-ver" style="color: var(--success-color); border-color: var(--success-color);">✅ v4.6.04 SHIPPED</span>
            <span class="phase-effort">~2h</span>
        </div>
        <div class="phase-title">Regression Coverage — 30+ Tests</div>
        <ul class="checklist">
            <li class="done">Audit all modules for their commands via browser_console</li>
            <li class="done">Add 2+ regression tests per module to <code>_regressionTasks()</code></li>
            <li class="done">Add "no regression tests" penalty to confidence score</li>
            <li class="done">Sync to <code>core-bundle.js</code></li>
            <li class="done">Browser verify: run full suite, confirm all pass</li>
        </ul>
    </div>

    <div class="phase-card">
        <div class="phase-header">
            <span class="phase-num">PHASE 3</span>
            <span class="phase-ver" style="color: var(--success-color); border-color: var(--success-color);">✅ v4.6.05 SHIPPED</span>
            <span class="phase-effort">~1.5h</span>
        </div>
        <div class="phase-title">Post-Commit Verification + Auto-Revert</div>
        <ul class="checklist">
            <li class="done">Add post-commit fetch + <code>new Function()</code> syntax check</li>
            <li class="done">Add auto-revert on syntax check failure (re-upload original code)</li>
            <li class="done">Add <code>last-good-version.json</code> tracking</li>
            <li class="done">Sync to <code>core-bundle.js</code></li>
            <li class="done">Browser verify: commit broken code, confirm auto-revert</li>
        </ul>
    </div>

    <div class="phase-card">
        <div class="phase-header">
            <span class="phase-num">PHASE 4</span>
            <span class="phase-ver" style="color: var(--success-color); border-color: var(--success-color);">✅ v4.6.06 SHIPPED</span>
            <span class="phase-effort">~2h</span>
        </div>
        <div class="phase-title">Confidence Gate + Auto-Approve</div>
        <ul class="checklist">
            <li class="done">Implement <code>_calculateConfidence()</code> scoring function</li>
            <li class="done">Add auto-approve logic to <code>ship_module</code> (score ≥ 70)</li>
            <li class="done">Add <code>auto-commit</code> flag to <code>commit_module</code></li>
            <li class="done">Add commit log with confidence scores</li>
            <li class="done">Sync to <code>core-bundle.js</code></li>
            <li class="done">Browser verify: high-confidence auto-commit works</li>
            <li class="done">Browser verify: low-confidence still requires human</li>
        </ul>
    </div>

    <div class="phase-card">
        <div class="phase-header">
            <span class="phase-num">PHASE 5</span>
            <span class="phase-ver" style="color: var(--success-color); border-color: var(--success-color);">✅ v4.6.08 SHIPPED</span>
            <span class="phase-effort">~1.5h</span>
        </div>
        <div class="phase-title">Self-Improvement Trigger — Autonomous Bug Detection</div>
        <ul class="checklist">
            <li class="done">Build <code>_toolToModuleMap</code> at init (map AI tools → source modules)</li>
            <li class="done">Extend <code>_postRunReview</code> with auto-fix suggestions</li>
            <li class="done">Add <code>_autoFixSuggestions</code> to system prompt context</li>
            <li class="done">Sync to <code>core-bundle.js</code></li>
            <li class="done">Browser verify: trigger tool failure, confirm suggestion appears</li>
        </ul>
    </div>

    <div class="phase-card">
        <div class="phase-header">
            <span class="phase-num">PHASE 6</span>
            <span class="phase-ver" style="color: var(--success-color); border-color: var(--success-color);">✅ v4.6.09 SHIPPED</span>
            <span class="phase-effort">~2.5h</span>
        </div>
        <div class="phase-title">Boot Watchdog — Recovery on Failed Boot</div>
        <ul class="checklist">
            <li class="done">Add boot-time health check in <code>main.html</code></li>
            <li class="done">Add auto-revert logic for failed boots</li>
            <li class="done">Add "safe mode" fallback (load last-known-good)</li>
            <li class="done">Browser verify: break a module, confirm boot watchdog reverts</li>
        </ul>
    </div>

    <div class="phase-card">
        <div class="phase-header">
            <span class="phase-num">PHASE 7</span>
            <span class="phase-ver" style="color: var(--warning-color); border-color: var(--warning-color);">🔶 v4.6.10 PLANNED</span>
            <span class="phase-effort">~4h</span>
        </div>
        <div class="phase-title">DebugSupervisorModule — Persistent Error DB + Clustering + Auto-Fix</div>
        <ul class="checklist">
            <li class="pending">Create <code>module-debug-supervisor.js</code> — core module, register via <code>webx11.moduleManager.registerModule()</code></li>
            <li class="pending">Error interception: console.error, window.onerror, unhandledrejection, fetch failures, tool errors</li>
            <li class="pending">Chain through DebugPanel interceptors (DebugPanel loads after core, must wrap supervisor's interceptors)</li>
            <li class="pending">VFS persistence: <code>/system/debug/errors.json</code> with dedup (Jaccard 0.35), LRU eviction, quota monitoring</li>
            <li class="pending">Error classification: type (ReferenceError/TypeError/HTTP/timeout), severity (fatal/error/warning), category (module/network/tool/security/vfs)</li>
            <li class="pending">Terminal commands: <code>debug-stats</code>, <code>debug-history</code>, <code>debug-analyze</code>, <code>debug-trends</code>, <code>debug-fix</code>, <code>debug-health</code></li>
            <li class="pending">Update all <code>_errorLog</code> consumers: WorldModelModule, virtual-fs.js /sys/ mounts, <code>_collectDashboardData()</code></li>
            <li class="pending">Wire <code>_autoFixSuggestions</code> into <code>buildSystemPrompt()</code> (activate dormant pipeline)</li>
            <li class="pending">Root-cause analysis: trace error chains across event loop turns with time-window correlation</li>
            <li class="pending">Cross-session learning: recurring patterns → auto-fix suggestions (confidence-gated)</li>
            <li class="pending">Self-diagnostic: <code>_healthy</code> flag, fallback to original <code>_errorLog</code> on init failure</li>
            <li class="pending">Add <code>debug-*</code> to <code>WebX11SecurityManager._baseWhitelist</code> in terminal.js</li>
            <li class="pending">Dashboard integration: replace inline error counting with supervisor data</li>
            <li class="pending">Docs: help.html, cheatsheet.html, data/news.json</li>
            <li class="pending">Browser verify: error capture, persistence, clustering, RCA, auto-fix, dashboard</li>
        </ul>
    </div>
</section>

<!-- ============ TESTING STRATEGY ============ -->
<section>
    <h2>🧪 Testing Strategy — How to Validate Without Human Interaction</h2>

    <h3>The Bootstrapping Problem</h3>
    <p style="color: var(--text-color); margin-bottom: 10px;">
        To test autonomous self-improvement without human interaction, we need to verify the system can:
        <strong style="color:var(--accent-color)">detect</strong> a bug,
        <strong style="color:var(--accent-color)">fix</strong> it,
        <strong style="color:var(--accent-color)">test</strong> the fix,
        <strong style="color:var(--accent-color)">commit</strong> without asking a human,
        <strong style="color:var(--accent-color)">verify</strong> the commit didn't break anything, and
        <strong style="color:var(--accent-color)">recover</strong> if it did.
    </p>

    <h3>Test Scenarios</h3>
    <table>
        <thead>
            <tr><th>ID</th><th>Scenario</th><th>Expected Behavior</th><th>Verification</th></tr>
        </thead>
        <tbody>
            <tr>
                <td class="warn">T1</td>
                <td>AI fixes a typo in CalcModule (1 line, non-critical)</td>
                <td class="pass">Confidence ≥ 85, auto-commit, post-commit syntax check passes</td>
                <td class="muted">Check <code>commit-log.json</code> for <code>autoApproved: true</code></td>
            </tr>
            <tr>
                <td class="warn">T2</td>
                <td>AI tries to rewrite ShellModule (100+ lines, critical)</td>
                <td class="warn">Confidence &lt; 70, requires human approval</td>
                <td class="muted">Pipeline output: <code>confidence: &lt;70, humanApprovalRequired: true</code></td>
            </tr>
            <tr>
                <td class="warn">T3</td>
                <td>AI edits a module with a <code>TypeError</code> in a method</td>
                <td class="fail">test_module catches JS error, auto-rollback</td>
                <td class="muted">test_module output: <code>errors: [...]</code></td>
            </tr>
            <tr>
                <td class="warn">T4</td>
                <td>AI commits code with syntax error that passes loadModuleFromCode</td>
                <td class="fail">Auto-revert fires, original code restored</td>
                <td class="muted">Commit log: <code>reverted: true, syntaxError: ...</code></td>
            </tr>
            <tr>
                <td class="warn">T5</td>
                <td>AI edits MindmapModule (0 regression tests)</td>
                <td class="warn">Confidence reduced by 30, likely requires human approval</td>
                <td class="muted">Check confidence score</td>
            </tr>
            <tr>
                <td class="warn">T6</td>
                <td>Manually break a module file on server, refresh page</td>
                <td class="pass">Boot watchdog detects failure, reverts to last-good</td>
                <td class="muted">Console: "Boot watchdog: reverting..."</td>
            </tr>
            <tr>
                <td class="warn">T7</td>
                <td>Trigger repeated <code>run_command</code> failures in a session</td>
                <td class="pass"><code>_postRunReview</code> suggests <code>ship_module</code> for broken module</td>
                <td class="muted">Check next AI turn for auto-fix suggestion</td>
            </tr>
            <tr>
                <td class="warn">T8</td>
                <td>Tell AI "the calc command is broken, fix it"</td>
                <td class="pass">Full autonomous cycle: edit → test → health → regress → score → auto-commit → verify</td>
                <td class="muted">Check commit log + server file + page refresh</td>
            </tr>
        </tbody>
    </table>

    <h3>Validation Procedure</h3>
    <div class="pipeline-box">
<pre>1. Open main.html in browser
2. Via browser_console, inject a known bug:
   <span class="step">ai.terminal.fs.writeFile('_mods/CalcModule.js', buggyCode)</span>
3. Tell the AI: "The calc command returns wrong results. Fix it."
4. AI should:
   a. Read the current CalcModule source
   b. Identify the bug
   c. <span class="step">Call ship_module({name:'CalcModule', code: fixedCode, tests:[{command:'calc 1+1', expect:'2'}], commit:true})</span>
   d. Pipeline runs: test → health → regression → score
   e. If score ≥ 70: auto-commit, post-commit verify
   f. If score &lt; 70: print "Human approval required, type commit-approve"
5. Verify:
   - <span class="ok">Commit log has entry</span>
   - <span class="ok">Server file updated (curl the .js URL)</span>
   - <span class="ok">Page refresh loads new version</span>
   - <span class="ok">calc 1+1 returns 2</span></pre>
    </div>
</section>

<!-- ============ RISKS ============ -->
<section>
    <h2>⚠️ Risks & Mitigations</h2>
    <table class="risk-matrix">
        <thead>
            <tr><th>Risk</th><th class="center">Severity</th><th>Mitigation</th></tr>
        </thead>
        <tbody>
            <tr>
                <td>AI commits broken code that passes tests</td>
                <td class="center risk-high">High</td>
                <td>Confidence score penalizes thin coverage; post-commit syntax check; boot watchdog</td>
            </tr>
            <tr>
                <td>Bundle sync corrupts core-bundle.js</td>
                <td class="center risk-high">High</td>
                <td><code>new Function()</code> validation before upload; auto-revert; git backup</td>
            </tr>
            <tr>
                <td>AI enters infinite fix loop</td>
                <td class="center risk-med">Medium</td>
                <td>Rate limit (5 commits/day); cooldown (10 min); max 3 retries per module per session</td>
            </tr>
            <tr>
                <td>Auto-fix suggestions pollute AI context</td>
                <td class="center risk-low">Low</td>
                <td>Cap suggestions at 3 per session; clear after each agentic loop</td>
            </tr>
            <tr>
                <td>Boot watchdog false positive</td>
                <td class="center risk-med">Medium</td>
                <td>Only trigger if module count &lt; 50% of expected; require 2 consecutive failed boots</td>
            </tr>
            <tr>
                <td>Confidence threshold too permissive</td>
                <td class="center risk-med">Medium</td>
                <td>Start at 70, tune based on commit success rate; log all auto-commits for review</td>
            </tr>
            <tr>
                <td>WebDAV token lacks write scope</td>
                <td class="center risk-low">Low</td>
                <td>Already verified working in v4.6.02 (TTS/SST changes were committed)</td>
            </tr>
        </tbody>
    </table>
</section>

<!-- ============ KEY ISSUES ============ -->
<section>
    <h2>🔬 Key Issues Found During Live Audit</h2>
    <table>
        <thead>
            <tr><th>#</th><th>Issue</th><th>Location</th><th>Evidence</th></tr>
        </thead>
        <tbody>
            <tr>
                <td class="fail">1</td>
                <td><code>_captureConsoleErrors</code> undefined — planned but never coded</td>
                <td class="muted">module-aiassist.js</td>
                <td class="muted"><code>typeof ai._captureConsoleErrors === 'undefined'</code></td>
            </tr>
            <tr>
                <td class="fail">2</td>
                <td><code>_rollbackModule</code> and <code>_runTestAssertion</code> not implemented as planned</td>
                <td class="muted">module-aiassist.js</td>
                <td class="muted">Both undefined; rollback is inline, not reusable</td>
            </tr>
            <tr>
                <td class="fail">3</td>
                <td>Health check doesn't check JS errors</td>
                <td class="muted">health_check handler</td>
                <td class="muted"><code>results.healthy = allCoreOk && versionMatch</code> — no jsErrors check</td>
            </tr>
            <tr>
                <td class="fail">4</td>
                <td>Regression suite returns false positive for untested modules</td>
                <td class="muted">_runRegressionSuite</td>
                <td class="muted"><code>total: 0 → allPassed: true</code> — passes by default</td>
            </tr>
            <tr>
                <td class="warn">5</td>
                <td><code>ship_module</code> with <code>commit:true</code> still creates pending commit (doesn't auto-execute)</td>
                <td class="muted">ship_module handler</td>
                <td class="muted">Calls commit_module → stores _pendingCommit — human gate always active</td>
            </tr>
            <tr>
                <td class="fail">6</td>
                <td>No <code>_toolToModuleMap</code> — Phase G can't map failures to modules</td>
                <td class="muted">module-aiassist.js</td>
                <td class="muted">No mapping between AI tool names and source modules</td>
            </tr>
        </tbody>
    </table>
</section>

<!-- ============ FILES ============ -->
<section>
    <h2>📁 Files Likely to Change</h2>
    <table>
        <thead>
            <tr><th>File</th><th>Changes</th><th>Phase</th></tr>
        </thead>
        <tbody>
            <tr>
                <td><code>module-aiassist.js</code></td>
                <td>_captureConsoleErrors, expanded _regressionTasks, _calculateConfidence, post-commit verify, _postRunReview extension, _toolToModuleMap</td>
                <td class="muted">1–5</td>
            </tr>
            <tr>
                <td><code>core-bundle.js</code></td>
                <td>Sync all changes from module-aiassist.js</td>
                <td class="muted">1–5</td>
            </tr>
            <tr>
                <td><code>main.html</code></td>
                <td>Boot watchdog script, version bump</td>
                <td class="muted">6</td>
            </tr>
            <tr>
                <td><code>module-loader.js</code></td>
                <td>Version bump</td>
                <td class="muted">6</td>
            </tr>
        </tbody>
    </table>
</section>

<!-- ============ OPEN QUESTIONS ============ -->
<section>
    <h2>❓ Open Questions</h2>
    <div class="open-q">
        <span class="q-num">Q1:</span>
        <span class="q-text">Confidence threshold starting point — 70 is my recommendation. Higher (80, more conservative) or lower (60, more permissive)? Can be tuned after observing auto-commit success rates.</span>
    </div>
    <div class="open-q">
        <span class="q-num">Q2:</span>
        <span class="q-text">Boot watchdog scope — auto-revert ANY failed boot, or only boots that fail within N seconds of a recent commit? A blanket auto-revert could undo intentional architecture changes.</span>
    </div>
    <div class="open-q">
        <span class="q-num">Q3:</span>
        <span class="q-text">Auto-fix suggestion delivery — inject into AI's system prompt (passive, safer) or auto-prompt the AI to fix them (active, risks runaway fix loops)?</span>
    </div>
    <div class="open-q">
        <span class="q-num">Q4:</span>
        <span class="q-text">Should <code>commit:true</code> in <code>ship_module</code> bypass the confidence gate? Options: (a) auto-commit if score ≥ threshold, (b) always human gate, (c) separate <code>auto:true</code> flag.</span>
    </div>
</section>

</div>

<footer>
    <p>newOS4 Self-Improving Autonomous OS Roadmap — Generated by Hermes Agent</p>
    <p>v4.6.03 LIVE · v4.6.09 SHIPPED · v4.6.10 PLANNED · Auto-commit active for confidence ≥ 70 · <a href="https://webx11.duckdns.org/spaces/newos4/main.html">Open newOS4 Terminal</a></p>
    <p style="margin-top: 8px; color: var(--success-color);">Last updated: 2026-07-07 · Status: 6/6 phases shipped, Phase 7 planned 🔶</p>
</footer>

</body>
</html>