HelixTranslate/docs/Fixed.html at main · HelixDevelopment/HelixTranslate · GitHub

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
<!DOCTYPE html>
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
  <meta charset="utf-8" />
  <meta name="generator" content="pandoc" />
  <meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=yes" />
  <title>Fixed</title>
  <style>
    /* Default styles provided by pandoc.
    ** See https://pandoc.org/MANUAL.html#variables-for-html for config info.
    */
    html {
      color: #1a1a1a;
      background-color: #fdfdfd;
    }
    body {
      margin: 0 auto;
      max-width: 36em;
      padding-left: 50px;
      padding-right: 50px;
      padding-top: 50px;
      padding-bottom: 50px;
      hyphens: auto;
      overflow-wrap: break-word;
      text-rendering: optimizeLegibility;
      font-kerning: normal;
    }
    @media (max-width: 600px) {
      body {
        font-size: 0.9em;
        padding: 12px;
      }
      h1 {
        font-size: 1.8em;
      }
    }
    @media print {
      html {
        background-color: white;
      }
      body {
        background-color: transparent;
        color: black;
        font-size: 12pt;
      }
      p, h2, h3 {
        orphans: 3;
        widows: 3;
      }
      h2, h3, h4 {
        page-break-after: avoid;
      }
    }
    p {
      margin: 1em 0;
    }
    a {
      color: #1a1a1a;
    }
    a:visited {
      color: #1a1a1a;
    }
    img {
      max-width: 100%;
    }
    svg {
      height: auto;
      max-width: 100%;
    }
    h1, h2, h3, h4, h5, h6 {
      margin-top: 1.4em;
    }
    h5, h6 {
      font-size: 1em;
      font-style: italic;
    }
    h6 {
      font-weight: normal;
    }
    ol, ul {
      padding-left: 1.7em;
      margin-top: 1em;
    }
    li > ol, li > ul {
      margin-top: 0;
    }
    blockquote {
      margin: 1em 0 1em 1.7em;
      padding-left: 1em;
      border-left: 2px solid #e6e6e6;
      color: #606060;
    }
    code {
      white-space: pre-wrap;
      font-family: Menlo, Monaco, Consolas, 'Lucida Console', monospace;
      font-size: 85%;
      margin: 0;
      hyphens: manual;
    }
    pre {
      margin: 1em 0;
      overflow: auto;
    }
    pre code {
      padding: 0;
      overflow: visible;
      overflow-wrap: normal;
    }
    .sourceCode {
     background-color: transparent;
     overflow: visible;
    }
    hr {
      border: none;
      border-top: 1px solid #1a1a1a;
      height: 1px;
      margin: 1em 0;
    }
    table {
      margin: 1em 0;
      border-collapse: collapse;
      width: 100%;
      overflow-x: auto;
      display: block;
      font-variant-numeric: lining-nums tabular-nums;
    }
    table caption {
      margin-bottom: 0.75em;
    }
    tbody {
      margin-top: 0.5em;
      border-top: 1px solid #1a1a1a;
      border-bottom: 1px solid #1a1a1a;
    }
    th {
      border-top: 1px solid #1a1a1a;
      padding: 0.25em 0.5em 0.25em 0.5em;
    }
    td {
      padding: 0.125em 0.5em 0.25em 0.5em;
    }
    header {
      margin-bottom: 4em;
      text-align: center;
    }
    #TOC li {
      list-style: none;
    }
    #TOC ul {
      padding-left: 1.3em;
    }
    #TOC > ul {
      padding-left: 0;
    }
    #TOC a:not(:hover) {
      text-decoration: none;
    }
    span.smallcaps{font-variant: small-caps;}
    div.columns{display: flex; gap: min(4vw, 1.5em);}
    div.column{flex: auto; overflow-x: auto;}
    div.hanging-indent{margin-left: 1.5em; text-indent: -1.5em;}
    /* The extra [class] is a hack that increases specificity enough to
       override a similar rule in reveal.js */
    ul.task-list[class]{list-style: none;}
    ul.task-list li input[type="checkbox"] {
      font-size: inherit;
      width: 0.8em;
      margin: 0 0.8em 0.2em -1.6em;
      vertical-align: middle;
    }
    .display.math{display: block; text-align: center; margin: 0.5rem auto;}
  </style>
</head>
<body>
<header id="title-block-header">
<h1 class="title">Fixed</h1>
</header>
<h1 id="helixtranslate--fixed-closed-workable-items">HelixTranslate —
Fixed (Closed Workable Items)</h1>
<p><strong>Revision:</strong> 1 <strong>Last modified:</strong>
2026-06-14T15:37:53Z <strong>Authority:</strong> §11.4.15 (status) ·
§11.4.16 (type) · §11.4.19 (column-alignment) · §11.4.33 (type-aware
closure vocabulary) · §11.4.53 (Fixed_Summary parity) · §11.4.54
(ATM-NNN ticket IDs) <strong>Scope:</strong> the closed-archive tracker.
Every entry is a real, landed, in-git fix from this session's
mutation-proven bug-hunt campaign + the format-matrix / PDF / DeepSeek /
WebSocket / SSH work. Each carries a stable <code>[ATM-NNN]</code> id,
<code>**Status:**</code>, <code>**Type:**</code>, a 1-2 line root cause,
and the commit hash as captured evidence (§11.4.123 — no fix is claimed
that is not in <code>git log</code>).</p>
<p>Sort order: ATM id ascending (allocation order). The companion
<code>Fixed_Summary.md</code> sorts most-recent-first.</p>
<hr />
<h3
id="1-atm-001-splittext-dropped-paragraph-separators-at-chunk-boundaries-causing-structural-data-loss">§1.
[ATM-001] splitText dropped paragraph separators at chunk boundaries
causing structural data loss</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: large-chapter chunking concatenated chunks without
re-inserting the paragraph separator at boundaries, silently flattening
structure in the translated output. Mutation-proven.</li>
<li>Evidence: commit <code>24b0fd0</code></li>
</ul>
<h3
id="2-atm-002-in-memory-translation-cache-key-collision-served-the-wrong-translation">§2.
[ATM-002] In-memory translation cache-key collision served the wrong
translation</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the in-memory cache key did not distinguish all inputs,
so distinct (text,context) tuples collided onto one entry and a prior
translation was returned for new content. Mutation-proven.</li>
<li>Evidence: commit <code>bd8a1ef</code></li>
</ul>
<h3
id="3-atm-003-fb2-writer-dropped-deeply-nested-2-level-section-content-as-translated-text-data-loss">§3.
[ATM-003] FB2 writer dropped deeply-nested (&gt;2 level) section content
as translated-text data loss</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the FB2 writer recursion stopped serializing section
content beyond two nesting levels, dropping translated text for deeply
structured books. Mutation-proven.</li>
<li>Evidence: commit <code>71babe1</code></li>
</ul>
<h3
id="4-atm-004-anthropic-client-dropped-all-content-blocks-after-the-first-truncating-translations">§4.
[ATM-004] Anthropic client dropped all content blocks after the first,
truncating translations</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the Anthropic response parser read only
<code>content[0]</code>, so multi-block responses were truncated or
emptied. Mutation-proven.</li>
<li>Evidence: commit <code>87a8048</code></li>
</ul>
<h3
id="5-atm-005-distributed-resultcache-spurious-eviction-on-update--round-robin-index-panic-after-pool-shrink">§5.
[ATM-005] Distributed ResultCache spurious eviction on update +
round-robin index panic after pool shrink</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: cache update path evicted a still-valid entry, and the
round-robin worker index was not re-bounded after the pool shrank,
causing an out-of-range panic. Mutation-proven.</li>
<li>Evidence: commit <code>edadd73</code></li>
</ul>
<h3
id="6-atm-006-epub-percent-encodedrelative-hrefs-dropped-chapters-and-cover-html-test-hack-corrupted-content">§6.
[ATM-006] EPUB percent-encoded/relative hrefs dropped chapters and
cover; HTML test-hack corrupted content</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the EPUB parser failed to resolve percent-encoded /
relative manifest hrefs (losing chapters + cover) and a leftover
hardcoded test-hack corrupted HTML content. Mutation-proven.</li>
<li>Evidence: commit <code>47ea69d</code></li>
</ul>
<h3
id="7-atm-007-verification-parsenote-dropped-examples-terminated-notes-produced-nan-confidence-under-counted-multipass-sections">§7.
[ATM-007] Verification parseNote dropped EXAMPLES-terminated notes,
produced NaN confidence, under-counted multipass sections</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: note parsing terminated early on EXAMPLES, divided by
zero into NaN confidence, and the multipass section counter was off.
Mutation-proven.</li>
<li>Evidence: commit <code>34e09f8</code></li>
</ul>
<h3
id="8-atm-008-redis-makecachekey-metadata-delimiter-injection-collision-served-the-wrong-translation">§8.
[ATM-008] Redis makeCacheKey metadata delimiter-injection collision
served the wrong translation</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the Redis cache key concatenated metadata with a
delimiter that could appear in the data, letting distinct inputs hash to
the same key and return a wrong cached translation.
Mutation-proven.</li>
<li>Evidence: commit <code>9946d85</code></li>
</ul>
<h3
id="9-atm-009-epubmarkdown-round-trip-destroyed-code-blocks-lost-gfm-tables-corrupted-backslash-escapes">§9.
[ATM-009] EPUB↔︎Markdown round-trip destroyed code blocks, lost GFM
tables, corrupted backslash-escapes</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the markdown round-trip converter mishandled fenced code
blocks, GFM tables, and backslash escapes, losing or corrupting content
on the EPUB→MD→EPUB path. Mutation-proven.</li>
<li>Evidence: commit <code>d055b8c</code></li>
</ul>
<h3
id="10-atm-010-api-translatetext-silently-dropped-scriptcyrillic-and-accepted-invalid-script-verifier-handlers-returned-wrong-http-status">§10.
[ATM-010] API translateText silently dropped script=cyrillic and
accepted invalid script; verifier handlers returned wrong HTTP
status</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the translate API ignored the cyrillic script flag and
did not validate the script value; verifier handlers mapped every
GetModel error to one wrong status code. Mutation-proven.</li>
<li>Evidence: commit <code>b8bc3ec</code></li>
</ul>
<h3
id="11-atm-011-verifier-fallback-chain-randomly-shuffled-lost-score-order-openrouter-pricing-decode-failed-on-string-values">§11.
[ATM-011] Verifier fallback chain randomly shuffled (lost score order);
OpenRouter pricing decode failed on string values</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the fallback chain was non-deterministically shuffled,
discarding score-priority order, and OpenRouter pricing decoding assumed
numeric JSON where the API returns strings. Mutation-proven.</li>
<li>Evidence: commit <code>4a95b06</code></li>
</ul>
<h3
id="12-atm-012-grpc-gettranslationstatus-returned-the-wrong-error-code-cleanupoldsessions-leaked-a-timeout-context">§12.
[ATM-012] gRPC GetTranslationStatus returned the wrong error code;
cleanupOldSessions leaked a timeout context</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: status lookup returned an incorrect gRPC code for
missing sessions, and the session-cleanup loop never cancelled its
per-iteration timeout context. Mutation-proven.</li>
<li>Evidence: commit <code>45f6152</code></li>
</ul>
<h3
id="13-atm-013-cwe-208-username-enumeration-timing-oracle-in-authenticateuser">§13.
[ATM-013] CWE-208 username-enumeration timing oracle in
AuthenticateUser</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: authentication short-circuited on unknown usernames
before the password hash compare, leaking account existence via response
timing. Mutation-proven.</li>
<li>Evidence: commit <code>5871848</code></li>
</ul>
<h3
id="14-atm-014-preparation-failed-chapter-analysis-misattribution--nested-subsection-analysis-content-loss">§14.
[ATM-014] Preparation failed-chapter analysis misattribution +
nested-subsection analysis content loss</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: a failed chapter analysis was attributed to the wrong
chapter index and nested subsection content was dropped from the
analysis pass. Mutation-proven.</li>
<li>Evidence: commit <code>367d88b</code></li>
</ul>
<h3
id="15-atm-015-directory-batch-to-a-non-existent-output-dir-collided-all-files-onto-one-path-data-loss">§15.
[ATM-015] Directory batch to a non-existent output dir collided all
files onto one path (data loss)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: batch directory translation did not create the output
directory and computed a single colliding output path, so every file
overwrote one another. Mutation-proven.</li>
<li>Evidence: commit <code>53b5056</code></li>
</ul>
<h3
id="16-atm-016-coordination-translatewithretry-ignored-ctx-cancellation-consensus-skipped-available-instances">§16.
[ATM-016] Coordination TranslateWithRetry ignored ctx cancellation;
consensus skipped available instances</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the retry loop did not honor context cancellation, and
the consensus selector skipped instances that were actually available.
Mutation-proven.</li>
<li>Evidence: commit <code>df8a8cb</code></li>
</ul>
<h3
id="17-atm-017-successful-but-empty-translation-silently-wiped-source-content-data-loss">§17.
[ATM-017] Successful-but-empty translation silently wiped source content
(data loss)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: when a provider returned an empty string with a success
status, the pipeline wrote the empty result over the source instead of
preserving the original. Mutation-proven.</li>
<li>Evidence: commit <code>837f528</code></li>
</ul>
<h3
id="18-atm-018-dashboard-eta-formatduration-trailing-space--non-unique-progress-event-ids">§18.
[ATM-018] Dashboard ETA formatDuration trailing space + non-unique
progress event IDs</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: <code>formatDuration</code> emitted a trailing space and
progress event IDs were not unique, breaking dashboard event
de-duplication. Mutation-proven.</li>
<li>Evidence: commit <code>b1b747f</code></li>
</ul>
<h3
id="19-atm-019-qwen-client-posted-to-native-dashscope-path-under-an-openai-compatible-base-wrong-url--response-shape-mismatch">§19.
[ATM-019] Qwen client posted to native DashScope path under an
OpenAI-compatible base (wrong URL + response-shape mismatch)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the Qwen client mixed the native DashScope endpoint with
an OpenAI-compatible base URL, producing a wrong request URL and a
response shape the parser could not read. Mutation-proven.</li>
<li>Evidence: commit <code>556002d</code></li>
</ul>
<h3
id="20-atm-020-preparation-translator--providers-flag-silently-ignored-cli-explicit--provider-overridden-by-config">§20.
[ATM-020] preparation-translator -providers flag silently ignored; CLI
explicit -provider overridden by config</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the <code>-providers</code> flag was parsed but never
consumed, and an explicit <code>-provider</code> was overwritten by the
config value. Mutation-proven.</li>
<li>Evidence: commit <code>5dfe003</code></li>
</ul>
<h3
id="21-atm-021-storage-cache-dup-tuple-served-stale-translation-added-unique-lookup-hash-index--idempotent-upsert">§21.
[ATM-021] Storage cache dup-tuple served stale translation (added UNIQUE
lookup-hash index + idempotent UPSERT)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: duplicate lookup tuples could store divergent rows with
no uniqueness constraint, serving a stale translation; fixed with a
UNIQUE index, idempotent UPSERT, and migration. Mutation-proven.</li>
<li>Evidence: commit <code>367adce</code></li>
</ul>
<h3
id="22-atm-022-versionmanagermetrics-data-race--lost-updates--11485-stresschaos-suites">§22.
[ATM-022] VersionManager.metrics data race / lost updates (+ §11.4.85
stress/chaos suites)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: concurrent metric updates in the distributed
VersionManager raced and lost increments; fixed with synchronization
plus stress/chaos coverage. Mutation-proven.</li>
<li>Evidence: commit <code>bf81517</code></li>
</ul>
<h3
id="23-atm-023-qwen-compatible-mode-alignment-to-default-base--11498-honest-live-test">§23.
[ATM-023] Qwen compatible-mode alignment to default base +
§11.4.98-honest live test</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: Qwen compatible-mode used an inconsistent base URL;
aligned to the default compatible base and made the live test honest per
§11.4.98.</li>
<li>Evidence: commit <code>2f0e359</code></li>
</ul>
<h3
id="24-atm-024-docx-input-non-functional-license-gated-unioffice--rewritten-as-a-stdlib-ooxml-parser">§24.
[ATM-024] DOCX input non-functional (license-gated unioffice) —
rewritten as a stdlib OOXML parser</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: DOCX parsing depended on the license-gated unioffice
library and was inert; rewritten as a stdlib OOXML parser, dropping the
dependency. Real-artifact proven.</li>
<li>Evidence: commit <code>d433210</code></li>
</ul>
<h3
id="25-atm-025-api-key-gate-ignored-provider-env-vars-deepseek_api_key-etc-in-unified-translator">§25.
[ATM-025] API-key gate ignored provider env vars (DEEPSEEK_API_KEY etc.)
in unified-translator</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the unified CLI API-key gate did not read
provider-specific environment variables, blocking otherwise-valid
providers. Real DeepSeek E2E proof committed.</li>
<li>Evidence: commit <code>c6c2930</code></li>
</ul>
<h3
id="26-atm-026-docx-rejected-by-the-cli-pipeline-despite-a-working-parser-detectorissupported-missing-docx">§26.
[ATM-026] DOCX rejected by the CLI pipeline despite a working parser
(detector.IsSupported missing DOCX)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the format detector's supported-set omitted DOCX, so the
CLI rejected files the parser could handle. Real E2E proven.</li>
<li>Evidence: commit <code>6304f60</code></li>
</ul>
<h3
id="27-atm-027-fb2--epub-input-broken-end-to-end-converttomarkdown-re-parsed-already-extracted-text">§27.
[ATM-027] FB2 + EPUB input broken end-to-end (convertToMarkdown
re-parsed already-extracted text)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the unified pipeline re-ran
<code>convertToMarkdown</code> over already-extracted FB2/EPUB text,
corrupting the content; fixed so all five input formats translate. Real
E2E matrix proven.</li>
<li>Evidence: commit <code>41063cb</code></li>
</ul>
<h3
id="28-atm-028-zhipu-missing-from-the-api-key-env-map-in-unified-translator--stale-allowlist-audit">§28.
[ATM-028] Zhipu missing from the API-key env map in unified-translator
(+ stale-allowlist audit)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the Zhipu provider was absent from the env-key map, so
its key was never picked up. Multi-provider E2E proof + stale-allowlist
audit committed.</li>
<li>Evidence: commit <code>9821e0f</code></li>
</ul>
<h3
id="29-atm-029-stale-provider-model-allowlists-blocked-current-geminigroqzhipu-models">§29.
[ATM-029] Stale provider model-allowlists blocked current
gemini/groq/zhipu models</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: hardcoded model allowlists predated current provider
catalogs, rejecting valid current models; added current models,
live-proven.</li>
<li>Evidence: commit <code>b23bcac</code></li>
</ul>
<h3
id="30-atm-030-coordination-consensus-tie-break-was-non-deterministic-map-iteration-order">§30.
[ATM-030] Coordination consensus tie-break was non-deterministic
(map-iteration order)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: consensus tie-breaking iterated a map, producing
non-deterministic winners across runs; made deterministic.
Mutation-proven.</li>
<li>Evidence: commit <code>77a0c15</code></li>
</ul>
<h3
id="31-atm-031-multipass-verification-chapter-loop-panicked-when-translation-has-fewer-chapters-than-original">§31.
[ATM-031] Multipass verification chapter loop panicked when translation
has fewer chapters than original</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the multipass loop indexed the translation by the
original's chapter count, panicking when the translation had fewer
chapters.</li>
<li>Evidence: commit <code>9640a00</code></li>
</ul>
<h3 id="32-atm-032-fb2-writer-dropped-sectionsubsection-titles">§32.
[ATM-032] FB2 writer dropped section/subsection titles</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the FB2 writer omitted section and subsection
<code>&lt;title&gt;</code> elements from the serialized output, losing
chapter headings. Mutation-proven.</li>
<li>Evidence: commit <code>b83038c</code></li>
</ul>
<h3
id="33-atm-033-pdf-input-revived--replaced-license-gated-unipdf-with-mit-ledongthucpdf">§33.
[ATM-033] PDF input revived — replaced license-gated unipdf with MIT
ledongthuc/pdf</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: PDF input was dead behind the license-gated unipdf
dependency; replaced with the MIT ledongthuc/pdf library.
Real-translation proven.</li>
<li>Evidence: commit <code>18d6f73</code></li>
</ul>
<h3
id="34-atm-034-cli-ignored--o-output-format--always-emitted-epub">§34.
[ATM-034] CLI ignored -o output format — always emitted EPUB</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the CLI hardcoded EPUB output regardless of the
<code>-o</code> extension; now honors <code>.txt/.fb2/.md/.epub</code>
(+ 2 §11.4.120 gate reconciliations).</li>
<li>Evidence: commit <code>ca5608f</code></li>
</ul>
<h3
id="35-atm-035-add-htmlhtm-cli-output-format-completes-the-5-output-matrix">§35.
[ATM-035] Add .html/.htm CLI output format (completes the 5-output
matrix)</h3>
<p><strong>Status:</strong> Implemented (→ Fixed.md)
<strong>Type:</strong> Feature</p>
<ul>
<li>Summary: added HTML-escaped <code>.html/.htm</code> output,
completing the output-format matrix. Real-translation proven.</li>
<li>Evidence: commit <code>fb07a59</code></li>
</ul>
<h3
id="36-atm-036-deepseek-model-allowlist-un-staled-accept-deepseek-v4-flashv4-pro">§36.
[ATM-036] DeepSeek model allowlist un-staled (accept
deepseek-v4-flash/v4-pro)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the DeepSeek allowlist predated v4 models and rejected
them; updated via live <code>/models</code> + real-translation
verification (the canonical allowlist-refresh template).</li>
<li>Evidence: commit <code>0fd1a34</code></li>
</ul>
<h3
id="37-atm-037-websocket-monitoring-test-harness-data-races--fatal-map-panic">§37.
[ATM-037] WebSocket monitoring test harness data races + fatal map
panic</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Task</p>
<ul>
<li>Root cause: the monitoring test harness had concurrent map access
causing data races and a fatal map-write panic; serialized access in the
harness.</li>
<li>Evidence: commit <code>ab7db0e</code></li>
</ul>
<h3
id="38-atm-038-testssherrorhandling-brittle-error-string-assertion-1141">§38.
[ATM-038] TestSSHErrorHandling brittle error-string assertion
(§11.4.1)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Task</p>
<ul>
<li>Root cause: the SSH error test asserted an exact error substring
that varied by environment, producing FAIL-bluffs; assert on a stable
condition instead.</li>
<li>Evidence: commit <code>817b9dd</code></li>
</ul>
<h3
id="39-atm-039-preparation-per-chapter-analysis-stamped-no-authoritative-chapter-number">§39.
[ATM-039] Preparation per-chapter analysis stamped no authoritative
chapter number</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: per-chapter analysis records lacked the authoritative
chapter number, breaking downstream attribution; now stamped.</li>
<li>Evidence: commit <code>0df25d9</code></li>
</ul>
<h3
id="40-atm-040-hardware-parselscpucores-undercounts-physical-cores-on-multi-socket-hosts">§40.
[ATM-040] hardware parseLscpuCores undercounts physical cores on
multi-socket hosts</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: <code>parseLscpuCores</code> counted cores per socket
once instead of multiplying across sockets, undercounting on
multi-socket machines and mis-tuning concurrency.</li>
<li>Evidence: commit <code>80d627b</code></li>
</ul>
<h3
id="41-atm-041-batch-same-stemdifferent-ext-inputs-collided-onto-one-output-file-data-loss">§41.
[ATM-041] Batch same-stem/different-ext inputs collided onto one output
file (data loss)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: batch output naming used only the file stem, so
<code>a.txt</code> and <code>a.fb2</code> produced the same output path
and overwrote each other.</li>
<li>Evidence: commit <code>c3117f8</code></li>
</ul>
<h3
id="42-atm-042-security-refreshtoken-resurrected-expiredunbounded-sessions-into-fresh-valid-tokens">§42.
[ATM-042] Security RefreshToken resurrected expired/unbounded sessions
into fresh valid tokens</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: token refresh did not re-check session expiry/bounds, so
an expired session could be exchanged for a fresh valid token.</li>
<li>Evidence: commit <code>d0fe40a</code></li>
</ul>
<h3
id="43-atm-043-emptynull-llmsverifier-envelope-modelsnull-failed-to-decode-breaking-apiv1verified-models">§43.
[ATM-043] Empty/null LLMsVerifier envelope ({"models":null}) failed to
decode, breaking /api/v1/verified-models</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the verified-models decoder did not tolerate a null
models array, so the endpoint broke whenever zero models were verified.
Mutation-proven.</li>
<li>Evidence: commit <code>d8142e5</code></li>
</ul>
<h3 id="44-atm-044-fb2--dropped-inline-element-text">§44. [ATM-044] FB2
<v>/<subtitle>/<text-author> dropped inline-element text</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the FB2 parser did not descend into inline elements
inside <code>&lt;v&gt;</code>, <code>&lt;subtitle&gt;</code>, and
<code>&lt;text-author&gt;</code>, dropping their text.
Mutation-proven.</li>
<li>Evidence: commit <code>8f52370</code></li>
</ul>
<h3
id="45-atm-045-distributed-fallback-getfailurerate-fabricated-a-100-rate-on-window-expiry--zero-window">§45.
[ATM-045] Distributed fallback getFailureRate fabricated a 100% rate on
window expiry / zero window</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the failure-rate calculation divided by an empty/expired
window and reported a fabricated 100% failure rate, triggering needless
fallback.</li>
<li>Evidence: commit <code>20beda7</code></li>
</ul>
<h3
id="46-atm-046-markdownepub-chapter--shipped-unescaped--malformed-xhtml">§46.
[ATM-046] markdown→EPUB chapter <title> shipped unescaped — malformed
XHTML</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the markdown→EPUB writer wrote chapter titles into XHTML
without escaping <code>&amp;</code>/<code>&lt;</code>/<code>&gt;</code>,
producing malformed XHTML for those titles. Mutation-proven.</li>
<li>Evidence: commit <code>be81550</code></li>
</ul>
<h3
id="47-atm-047-verification-polishchapterpolishsectionrecursive-index-out-of-range-on-shorter-translations">§47.
[ATM-047] Verification polishChapter/polishSectionRecursive
index-out-of-range on shorter translations</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the polish pass indexed by the original
section/subsection count and went out of range when the translation had
fewer. Mutation-proven.</li>
<li>Evidence: commit <code>7434ee4</code></li>
</ul>
<h3
id="48-atm-048-ollama-client-dropped-temperaturemax_tokens--never-sent-options-to-apigenerate">§48.
[ATM-048] Ollama client dropped temperature/max_tokens — never sent
options to /api/generate</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the Ollama client never populated the
<code>options</code> field, so temperature and max_tokens were silently
ignored. Mutation-proven.</li>
<li>Evidence: commit <code>a728e57</code></li>
</ul>
<h3
id="49-atm-049-epub-chapter-text-shipped-html-entities-literally-to-readers">§49.
[ATM-049] EPUB chapter text shipped HTML entities literally to
readers</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: EPUB chapter text was emitted with HTML entities
un-decoded, so readers saw literal entity sequences.
Mutation-proven.</li>
<li>Evidence: commit <code>107e570</code></li>
</ul>
<h3
id="50-atm-050-grpc-subscribeevents-delivered-lifecycle-events-with-an-empty-session_id">§50.
[ATM-050] gRPC SubscribeEvents delivered lifecycle events with an empty
session_id</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the SubscribeEvents stream forwarded lifecycle events
without populating <code>session_id</code>, breaking client-side session
filtering.</li>
<li>Evidence: commit <code>fd07e18</code></li>
</ul>
<h3
id="51-atm-051-loadconfig-nil-map-panic-when-config-omits-translationproviders-and-a-provider-key-env-var-is-set">§51.
[ATM-051] LoadConfig nil-map panic when config omits
translation.providers and a provider key env var is set</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: <code>LoadConfig</code> wrote into a nil providers map
when the config omitted <code>translation.providers</code> but a
provider API-key env var was present, panicking on startup.</li>
<li>Evidence: commit <code>fa35d2e</code></li>
</ul>
<h3
id="52-atm-052-epub-cover-image-mislabeled-imagejpeg-on-round-trip-broke-pnggifwebpsvg-covers">§52.
[ATM-052] EPUB cover image mislabeled image/jpeg on round-trip (broke
PNG/GIF/WEBP/SVG covers)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the EPUB writer hardcoded <code>image/jpeg</code> for
the cover media type, breaking non-JPEG covers on round-trip.</li>
<li>Evidence: commit <code>c416bd8</code></li>
</ul>
<h3
id="53-atm-053-basetranslator-cachestats-data-race--concurrent-map-write-crash">§53.
[ATM-053] BaseTranslator cache/stats data race + concurrent-map-write
crash</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: concurrent translators accessed the cache/stats maps
without synchronization, causing data races and a concurrent-map-write
crash. Race-proven.</li>
<li>Evidence: commit <code>8316e3f</code></li>
</ul>
<h3
id="54-atm-054-markdown-translator--format-flag-unvalidated--unsupported-value-produced-no-output-yet-reported-success">§54.
[ATM-054] markdown-translator -format flag unvalidated — unsupported
value produced no output yet reported success</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the markdown-translator did not validate
<code>-format</code> against <code>{epub,md}</code>, so an unsupported
value silently produced no output while reporting success.</li>
<li>Evidence: commit <code>85b3362</code></li>
</ul>
<h3
id="55-atm-055-cli-ssh-provider-ignored--sourcetarget-lang--script--llama-binary-model-hardcoded-rusr-cyrillic">§55.
[ATM-055] CLI SSH provider ignored -source/target-lang, -script,
-llama-binary/-model (hardcoded ru→sr-cyrillic)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the SSH provider path hardcoded ru→sr-cyrillic and
ignored the language/script/llama flags, so non-default jobs produced
wrong output.</li>
<li>Evidence: commit <code>36f740a</code></li>
</ul>
<h3
id="56-atm-056-cmdapi-server-source-was-gitignored-committed-health-check-regression-guard">§56.
[ATM-056] cmd/api-server source was gitignored; committed health-check
regression guard</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Task</p>
<ul>
<li>Root cause: a <code>.gitignore</code> rule excluded
<code>cmd/api-server/</code> source (§11.4.30 violation); un-ignored it
and committed a health-check regression guard.</li>
<li>Evidence: commit <code>4bbac9a</code></li>
</ul>
<h3
id="57-atm-057-cmdcli-gemini_api_key-silently-ignored--gemini-missing-from-env-key-map">§57.
[ATM-057] cmd/cli GEMINI_API_KEY silently ignored — gemini missing from
env-key map</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the <code>cmd/cli</code> env-key map omitted gemini, so
<code>GEMINI_API_KEY</code> was never read.</li>
<li>Evidence: commit <code>309ca91</code></li>
</ul>
<h3
id="58-atm-058-translate-ssh-delivered-the-translated-epub-to-the-wrong-path--output-ignored">§58.
[ATM-058] translate-ssh delivered the translated EPUB to the wrong path;
-output ignored</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the SSH worker wrote the translated EPUB to a derived
path and ignored <code>-output</code>, delivering the result to the
wrong location.</li>
<li>Evidence: commit <code>86d10d4</code></li>
</ul>
<h3
id="59-atm-059-grpc-server-ignored-documented-env-var-overrides-grpc_addressport-log_level-enable_metricsreflection">§59.
[ATM-059] grpc-server ignored documented env-var overrides
(GRPC_ADDRESS/PORT, LOG_LEVEL, ENABLE_METRICS/REFLECTION)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the gRPC server documented env-var overrides but never
read them, so configuration via environment had no effect.</li>
<li>Evidence: commit <code>21c2460</code></li>
</ul>
<h3
id="60-atm-060-cmddeployment-status-action-panicked-on-container-ids-shorter-than-12-chars">§60.
[ATM-060] cmd/deployment status action panicked on container IDs shorter
than 12 chars</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the deployment status formatter sliced the container ID
to 12 chars without a length check, panicking on shorter IDs.</li>
<li>Evidence: commit <code>cde2e1a</code></li>
</ul>
<h3
id="61-atm-061-grpc-servermonitor-server-dead--max-connections-flag-wired--monitor-start-error-no-longer-swallowed">§61.
[ATM-061] grpc-server/monitor-server: dead -max-connections flag wired +
monitor start error no longer swallowed</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the <code>-max-connections</code> flag was parsed but
inert, and the monitor server's start error was discarded, hiding
startup failures. Mutation-proven.</li>
<li>Evidence: commit <code>93e89d7</code></li>
</ul>
<h3
id="62-atm-062-logger-json-reserved-key-collision-silently-dropped-real-severitymessagetimestamp">§62.
[ATM-062] logger JSON reserved-key collision silently dropped real
severity/message/timestamp</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: structured log fields named like reserved keys
(severity/message/timestamp) overwrote the real log values in the JSON
output. Mutation-proven.</li>
<li>Evidence: commit <code>daa2f70</code></li>
</ul>
<h3
id="63-atm-063-enhancetranslation-capitalization-dead-for-cyrillicaccented-latin-first-letters-byte-indexed-guard">§63.
[ATM-063] enhanceTranslation capitalization dead for
Cyrillic/accented-Latin first letters (byte-indexed guard)</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Bug</p>
<ul>
<li>Root cause: the capitalization guard used a byte index assuming
ASCII, so multi-byte Cyrillic/accented-Latin first letters were never
capitalized. Mutation-proven.</li>
<li>Evidence: commit <code>e520c4c</code></li>
</ul>
<h3
id="64-atm-064-single-source-binary-version-from-authoritative-version">§64.
[ATM-064] Single-source binary version from authoritative VERSION</h3>
<p><strong>Status:</strong> Fixed (→ Fixed.md) <strong>Type:</strong>
Task</p>
<ul>
<li>Root cause: binaries hardcoded divergent <code>appVersion</code>
literals; reconciled to read the single authoritative
<code>VERSION</code> (P0.1). Mutation-proven.</li>
<li>Evidence: commit <code>a36030e</code></li>
</ul>
<hr />
<p><em>This archive is regenerated into <code>Fixed_Summary.md</code> by
<code>scripts/testing/generate_fixed_summary.sh</code>. Do not hand-edit
the summary.</em></p>
</body>
</html>