-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
877 lines (784 loc) · 26.9 KB
/
index.html
File metadata and controls
877 lines (784 loc) · 26.9 KB
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
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Skill Forge v2 — Autonomous AI Skill & Code Improvement</title>
<meta name="description" content="An autonomous experiment loop that makes AI skills and codebases better while you sleep. Two modes: Skill optimization and Generic metric optimization. Auto or Guided execution.">
<style>
:root {
--bg: #0a0a0f;
--surface: #12121a;
--border: #1e1e2e;
--text: #e0e0e8;
--muted: #8888a0;
--accent: #f59e0b;
--accent2: #3b82f6;
--green: #10b981;
--glow: rgba(245, 158, 11, 0.15);
}
* { margin: 0; padding: 0; box-sizing: border-box; }
body {
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', system-ui, sans-serif;
background: var(--bg);
color: var(--text);
line-height: 1.7;
overflow-x: hidden;
}
.grid-bg {
position: fixed;
inset: 0;
background-image:
linear-gradient(rgba(245, 158, 11, 0.03) 1px, transparent 1px),
linear-gradient(90deg, rgba(245, 158, 11, 0.03) 1px, transparent 1px);
background-size: 60px 60px;
pointer-events: none;
z-index: 0;
}
.container {
max-width: 960px;
margin: 0 auto;
padding: 0 24px;
position: relative;
z-index: 1;
}
/* Hero */
.hero {
padding: 100px 0 80px;
text-align: center;
}
.hero-badge {
display: inline-block;
padding: 6px 16px;
border: 1px solid var(--border);
border-radius: 999px;
font-size: 13px;
color: var(--muted);
margin-bottom: 32px;
letter-spacing: 0.5px;
}
.hero-badge span { color: var(--accent); }
h1 {
font-size: clamp(2.5rem, 6vw, 4rem);
font-weight: 800;
letter-spacing: -0.03em;
line-height: 1.1;
margin-bottom: 24px;
}
h1 .forge {
background: linear-gradient(135deg, var(--accent), #ef4444);
-webkit-background-clip: text;
-webkit-text-fill-color: transparent;
background-clip: text;
}
.hero p {
font-size: 1.2rem;
color: var(--muted);
max-width: 640px;
margin: 0 auto 40px;
}
.cta-row {
display: flex;
gap: 16px;
justify-content: center;
flex-wrap: wrap;
}
.btn {
display: inline-flex;
align-items: center;
gap: 8px;
padding: 14px 28px;
border-radius: 10px;
font-size: 15px;
font-weight: 600;
text-decoration: none;
transition: all 0.2s;
}
.btn-primary {
background: var(--accent);
color: #000;
}
.btn-primary:hover {
transform: translateY(-2px);
box-shadow: 0 8px 30px var(--glow);
}
.btn-secondary {
border: 1px solid var(--border);
color: var(--text);
background: var(--surface);
}
.btn-secondary:hover {
border-color: var(--accent);
color: var(--accent);
}
/* Loop animation */
.loop-visual {
margin: 60px auto 0;
max-width: 700px;
background: var(--surface);
border: 1px solid var(--border);
border-radius: 16px;
padding: 40px 32px;
position: relative;
overflow: hidden;
}
.loop-visual::before {
content: '';
position: absolute;
top: -1px;
left: 0;
right: 0;
height: 2px;
background: linear-gradient(90deg, transparent, var(--accent), transparent);
animation: scan 3s ease-in-out infinite;
}
@keyframes scan {
0%, 100% { opacity: 0.3; }
50% { opacity: 1; }
}
.loop-steps {
display: flex;
align-items: center;
justify-content: center;
flex-wrap: wrap;
gap: 8px;
font-family: 'SF Mono', 'Fira Code', monospace;
font-size: 14px;
}
.loop-step {
padding: 8px 14px;
background: rgba(245, 158, 11, 0.08);
border: 1px solid rgba(245, 158, 11, 0.2);
border-radius: 8px;
color: var(--accent);
white-space: nowrap;
}
.loop-arrow { color: var(--muted); font-size: 18px; }
.loop-label {
text-align: center;
margin-top: 16px;
font-size: 13px;
color: var(--muted);
}
/* Sections */
section { padding: 80px 0; }
h2 {
font-size: 2rem;
font-weight: 700;
margin-bottom: 16px;
letter-spacing: -0.02em;
}
h2 .highlight { color: var(--accent); }
h2 .highlight-blue { color: var(--accent2); }
.section-sub {
color: var(--muted);
font-size: 1.05rem;
margin-bottom: 48px;
max-width: 640px;
}
/* Mode cards */
.mode-grid {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(280px, 1fr));
gap: 20px;
margin-bottom: 48px;
}
.mode-card {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 14px;
padding: 28px;
transition: border-color 0.2s;
}
.mode-card:hover { border-color: var(--accent); }
.mode-card h3 {
font-size: 1.15rem;
margin-bottom: 8px;
}
.mode-card .tag {
display: inline-block;
padding: 3px 10px;
border-radius: 6px;
font-size: 11px;
font-weight: 600;
text-transform: uppercase;
letter-spacing: 0.5px;
margin-bottom: 12px;
}
.tag-skill { background: rgba(245, 158, 11, 0.15); color: var(--accent); }
.tag-generic { background: rgba(59, 130, 246, 0.15); color: var(--accent2); }
.tag-auto { background: rgba(16, 185, 129, 0.15); color: var(--green); }
.tag-guided { background: rgba(168, 85, 247, 0.15); color: #a855f7; }
.mode-card p {
color: var(--muted);
font-size: 14px;
line-height: 1.6;
}
.mode-card ul {
list-style: none;
margin-top: 12px;
}
.mode-card ul li {
color: var(--muted);
font-size: 13px;
padding: 3px 0;
}
.mode-card ul li::before {
content: '\2192';
margin-right: 8px;
color: var(--accent);
}
/* Agent cards */
.agent-grid {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(260px, 1fr));
gap: 20px;
}
.agent-card {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 14px;
padding: 28px;
transition: border-color 0.2s;
}
.agent-card:hover { border-color: var(--accent); }
.agent-icon {
width: 44px;
height: 44px;
border-radius: 10px;
display: flex;
align-items: center;
justify-content: center;
font-size: 22px;
margin-bottom: 16px;
}
.agent-card:nth-child(1) .agent-icon { background: rgba(245, 158, 11, 0.12); }
.agent-card:nth-child(2) .agent-icon { background: rgba(59, 130, 246, 0.12); }
.agent-card:nth-child(3) .agent-icon { background: rgba(16, 185, 129, 0.12); }
.agent-card h3 { font-size: 1.1rem; margin-bottom: 4px; }
.agent-card .role { color: var(--muted); font-size: 13px; margin-bottom: 12px; }
.agent-card p { color: var(--muted); font-size: 14px; line-height: 1.6; }
/* Score formula */
.formula-box {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 14px;
padding: 32px;
margin: 40px 0;
font-family: 'SF Mono', 'Fira Code', monospace;
text-align: center;
}
.formula-label {
font-family: -apple-system, sans-serif;
font-size: 13px;
color: var(--muted);
text-transform: uppercase;
letter-spacing: 1px;
margin-bottom: 16px;
}
.formula { font-size: 1.1rem; color: var(--accent); }
/* Results */
.results-grid {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(280px, 1fr));
gap: 20px;
}
.result-card {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 14px;
padding: 28px;
}
.result-card h3 { font-size: 1rem; margin-bottom: 6px; }
.result-card .desc { color: var(--muted); font-size: 13px; margin-bottom: 20px; }
.score-bar {
display: flex;
align-items: center;
gap: 12px;
margin-bottom: 10px;
}
.score-label { font-size: 13px; color: var(--muted); width: 50px; }
.bar-track {
flex: 1;
height: 8px;
background: rgba(255,255,255,0.05);
border-radius: 4px;
overflow: hidden;
}
.bar-fill { height: 100%; border-radius: 4px; transition: width 1.5s ease; }
.bar-fill.before { background: var(--muted); }
.bar-fill.after { background: var(--accent); }
.score-val {
font-family: 'SF Mono', monospace;
font-size: 13px;
width: 42px;
text-align: right;
}
.result-insight {
margin-top: 16px;
padding-top: 16px;
border-top: 1px solid var(--border);
font-size: 13px;
color: var(--muted);
font-style: italic;
}
/* Features */
.feature-list {
display: grid;
grid-template-columns: repeat(auto-fit, minmax(260px, 1fr));
gap: 16px;
}
.feature {
display: flex;
gap: 14px;
padding: 20px;
background: var(--surface);
border: 1px solid var(--border);
border-radius: 12px;
}
.feature-dot {
width: 8px;
height: 8px;
border-radius: 50%;
background: var(--accent);
margin-top: 8px;
flex-shrink: 0;
}
.feature h4 { font-size: 0.95rem; margin-bottom: 4px; }
.feature p { font-size: 13px; color: var(--muted); line-height: 1.5; }
/* Code block */
.code-section {
background: var(--surface);
border: 1px solid var(--border);
border-radius: 14px;
overflow: hidden;
margin: 40px 0;
}
.code-header {
padding: 12px 20px;
border-bottom: 1px solid var(--border);
font-size: 13px;
color: var(--muted);
display: flex;
align-items: center;
gap: 8px;
}
.code-header .dot { width: 10px; height: 10px; border-radius: 50%; }
.code-header .dot:nth-child(1) { background: #ef4444; }
.code-header .dot:nth-child(2) { background: #f59e0b; }
.code-header .dot:nth-child(3) { background: #10b981; }
.code-body {
padding: 20px 24px;
font-family: 'SF Mono', 'Fira Code', monospace;
font-size: 14px;
line-height: 1.8;
overflow-x: auto;
color: var(--muted);
}
.code-body .comment { color: #555570; }
.code-body .cmd { color: var(--accent); }
.code-body .string { color: #10b981; }
/* Guided steps */
.guided-steps {
display: flex;
flex-direction: column;
gap: 12px;
margin-top: 24px;
}
.guided-step {
display: flex;
align-items: center;
gap: 16px;
padding: 16px 20px;
background: var(--surface);
border: 1px solid var(--border);
border-radius: 10px;
}
.step-num {
width: 32px;
height: 32px;
border-radius: 8px;
background: rgba(168, 85, 247, 0.12);
color: #a855f7;
display: flex;
align-items: center;
justify-content: center;
font-weight: 700;
font-size: 14px;
flex-shrink: 0;
}
.step-content h4 { font-size: 0.9rem; margin-bottom: 2px; }
.step-content p { font-size: 13px; color: var(--muted); }
/* Footer */
footer {
padding: 60px 0 40px;
border-top: 1px solid var(--border);
text-align: center;
}
footer p { color: var(--muted); font-size: 14px; line-height: 1.8; }
footer a { color: var(--accent); text-decoration: none; }
footer a:hover { text-decoration: underline; }
.footer-links {
display: flex;
justify-content: center;
gap: 24px;
margin-bottom: 20px;
}
@media (max-width: 640px) {
.hero { padding: 60px 0 40px; }
section { padding: 50px 0; }
.loop-visual { padding: 24px 16px; }
.loop-steps { gap: 6px; }
.loop-step { padding: 6px 10px; font-size: 12px; }
}
</style>
</head>
<body>
<div class="grid-bg"></div>
<!-- Hero -->
<header class="hero">
<div class="container">
<div class="hero-badge">v2 <span>·</span> Open Source <span>·</span> MIT License <span>·</span> Claude Cowork <span>·</span> OpenClaw</div>
<h1>Skill <span class="forge">Forge</span></h1>
<p>An autonomous experiment loop that makes AI skills and codebases better while you sleep. Works with Claude Cowork and OpenClaw. Run it fully autonomous or step through each decision in guided mode.</p>
<div class="cta-row">
<a href="https://github.com/GodModeAI2025/skill-forge" class="btn btn-primary">
<svg width="18" height="18" viewBox="0 0 16 16" fill="currentColor"><path d="M8 0C3.58 0 0 3.58 0 8c0 3.54 2.29 6.53 5.47 7.59.4.07.55-.17.55-.38 0-.19-.01-.82-.01-1.49-2.01.37-2.53-.49-2.69-.94-.09-.23-.48-.94-.82-1.13-.28-.15-.68-.52-.01-.53.63-.01 1.08.58 1.23.82.72 1.21 1.87.87 2.33.66.07-.52.28-.87.51-1.07-1.78-.2-3.64-.89-3.64-3.95 0-.87.31-1.59.82-2.15-.08-.2-.36-1.02.08-2.12 0 0 .67-.21 2.2.82.64-.18 1.32-.27 2-.27.68 0 1.36.09 2 .27 1.53-1.04 2.2-.82 2.2-.82.44 1.1.16 1.92.08 2.12.51.56.82 1.27.82 2.15 0 3.07-1.87 3.75-3.65 3.95.29.25.54.73.54 1.48 0 1.07-.01 1.93-.01 2.2 0 .21.15.46.55.38A8.013 8.013 0 0016 8c0-4.42-3.58-8-8-8z"/></svg>
View on GitHub
</a>
<a href="#quickstart" class="btn btn-secondary">Quick Start</a>
<a href="https://godmodeai2025.github.io/skill-forge/index_de.html" class="btn btn-primary">Schulung</a>
</div>
<div class="loop-visual">
<div class="loop-steps">
<span class="loop-step">Wizard</span>
<span class="loop-arrow">→</span>
<span class="loop-step">Dry-Run</span>
<span class="loop-arrow">→</span>
<span class="loop-step">Hypothesize</span>
<span class="loop-arrow">→</span>
<span class="loop-step">Mutate</span>
<span class="loop-arrow">→</span>
<span class="loop-step">Evaluate</span>
<span class="loop-arrow">→</span>
<span class="loop-step">Score</span>
<span class="loop-arrow">→</span>
<span class="loop-step">Keep / Revert</span>
<span class="loop-arrow">↺</span>
</div>
<div class="loop-label">Setup Wizard with Validation Gates → Experiment Loop → Morning Report</div>
</div>
</div>
</header>
<!-- Two Modes -->
<section>
<div class="container">
<h2>Two <span class="highlight">Domains</span>, Two <span class="highlight-blue">Execution</span> Modes</h2>
<p class="section-sub">Optimize SKILL.md files or any codebase. Run fully autonomous or step through each decision.</p>
<div class="mode-grid">
<div class="mode-card">
<span class="tag tag-skill">Skill Mode</span>
<h3>Optimize AI Skills</h3>
<p>Improve a Skill's SKILL.md by running evals, grading outputs, and scoring with a composite metric. Works with Claude Cowork and OpenClaw skills.</p>
<ul>
<li>Assertion-based scoring</li>
<li>Optional LLM-as-Judge</li>
<li>Train/test split for overfitting protection</li>
</ul>
</div>
<div class="mode-card">
<span class="tag tag-generic">Generic Mode</span>
<h3>Optimize Any Metric</h3>
<p>Improve any file against any shell command that returns a number: test coverage, bundle size, Lighthouse, Docker image size, lint errors.</p>
<ul>
<li>Any shell command as metric</li>
<li>Higher-is-better or lower-is-better</li>
<li>Glob pattern for scope</li>
</ul>
</div>
<div class="mode-card">
<span class="tag tag-auto">Auto Mode</span>
<h3>Run Unattended</h3>
<p>Fully autonomous loop. Start it, go to sleep, wake up to a morning report with measurable improvements.</p>
<ul>
<li>Perfect for Scheduled Tasks</li>
<li>Automatic keep/revert decisions</li>
<li>No user interaction needed</li>
</ul>
</div>
<div class="mode-card">
<span class="tag tag-guided">Guided Mode</span>
<h3>Decide at Every Step</h3>
<p>Interactive loop with 5 checkpoints where you review, adjust, and approve each experiment before it runs.</p>
<ul>
<li>Review and edit evals</li>
<li>Approve or redirect hypotheses</li>
<li>Override keep/revert decisions</li>
</ul>
</div>
</div>
</div>
</section>
<!-- Guided Checkpoints -->
<section>
<div class="container">
<h2>Guided Mode: <span class="highlight">5 Checkpoints</span></h2>
<p class="section-sub">In guided mode, you stay in control while the agents do the heavy lifting.</p>
<div class="guided-steps">
<div class="guided-step">
<div class="step-num">1</div>
<div class="step-content">
<h4>Review Evals</h4>
<p>See generated test cases, adjust assertions, change weights, add your own edge cases.</p>
</div>
</div>
<div class="guided-step">
<div class="step-num">2</div>
<div class="step-content">
<h4>Approve Hypothesis</h4>
<p>The agent proposes what to change and why. Accept, modify, or provide your own direction.</p>
</div>
</div>
<div class="guided-step">
<div class="step-num">3</div>
<div class="step-content">
<h4>Review Mutation Diff</h4>
<p>See exactly what will change before it's applied. Edit, approve, or skip.</p>
</div>
</div>
<div class="guided-step">
<div class="step-num">4</div>
<div class="step-content">
<h4>Evaluate Results</h4>
<p>See the score before and after. Override the automatic recommendation if you know better.</p>
</div>
</div>
<div class="guided-step">
<div class="step-num">5</div>
<div class="step-content">
<h4>Continue or Stop</h4>
<p>See progress so far and decide: another round, N more rounds, or generate the final report.</p>
</div>
</div>
</div>
</div>
</section>
<!-- Agents -->
<section>
<div class="container">
<h2>Three <span class="highlight">Agents</span>, One Loop</h2>
<p class="section-sub">Each experiment cycle is driven by three specialized agents working in sequence.</p>
<div class="agent-grid">
<div class="agent-card">
<div class="agent-icon">🔬</div>
<h3>Hypothesis Agent</h3>
<div class="role">The Scientist</div>
<p>Analyzes failures, consults the coverage matrix, and produces a testable hypothesis targeting the highest-impact root cause.</p>
</div>
<div class="agent-card">
<div class="agent-icon">🩹</div>
<h3>Mutator Agent</h3>
<div class="role">The Surgeon</div>
<p>Applies one minimal, focused change. Works on SKILL.md instructions or code files. Documents every change with category tags.</p>
</div>
<div class="agent-card">
<div class="agent-icon">⚖</div>
<h3>Scorer Agent</h3>
<div class="role">The Judge</div>
<p>LLM-as-Judge evaluating output quality across four dimensions. In Generic Mode, the shell command provides the score directly.</p>
</div>
</div>
</div>
</section>
<!-- Scoring -->
<section>
<div class="container">
<h2>Objective <span class="highlight">Scoring</span></h2>
<p class="section-sub">Every mutation is measured against a composite metric. No vibes, no gut feeling — just numbers.</p>
<div class="formula-box">
<div class="formula-label">Skill Mode — Composite Score</div>
<div class="formula">assertion_pass_rate × 0.80 + efficiency_score × 0.20</div>
</div>
<div class="formula-box">
<div class="formula-label">Skill Mode — With LLM-as-Judge</div>
<div class="formula">assertions × 0.50 + llm_judge × 0.30 + efficiency × 0.20</div>
</div>
<div class="formula-box">
<div class="formula-label">Generic Mode — Direct Metric</div>
<div class="formula">shell_command → extract_number → compare_to_baseline → keep/revert</div>
</div>
</div>
</section>
<!-- Results -->
<section>
<div class="container">
<h2>Real-World <span class="highlight">Results</span></h2>
<p class="section-sub">Tested on production skills with measurable before/after improvements.</p>
<div class="results-grid">
<div class="result-card">
<h3>humanizer</h3>
<div class="desc">Text humanization · 3 experiments</div>
<div class="score-bar">
<span class="score-label">Before</span>
<div class="bar-track"><div class="bar-fill before" style="width:74%"></div></div>
<span class="score-val">0.74</span>
</div>
<div class="score-bar">
<span class="score-label">After</span>
<div class="bar-track"><div class="bar-fill after" style="width:90%"></div></div>
<span class="score-val">0.90</span>
</div>
<div class="result-insight">+21.6% — Personality as a dedicated workflow step with concrete criteria.</div>
</div>
<div class="result-card">
<h3>fachbuch-lektorat</h3>
<div class="desc">German technical book editing · 3 experiments</div>
<div class="score-bar">
<span class="score-label">Before</span>
<div class="bar-track"><div class="bar-fill before" style="width:87%"></div></div>
<span class="score-val">87%</span>
</div>
<div class="score-bar">
<span class="score-label">After</span>
<div class="bar-track"><div class="bar-fill after" style="width:100%"></div></div>
<span class="score-val">100%</span>
</div>
<div class="result-insight">Worked examples beat abstract rules for edge-case handling.</div>
</div>
<div class="result-card">
<h3>was-bisher-geschah</h3>
<div class="desc">AI news briefing · 1 experiment</div>
<div class="score-bar">
<span class="score-label">Before</span>
<div class="bar-track"><div class="bar-fill before" style="width:93%"></div></div>
<span class="score-val">93%</span>
</div>
<div class="score-bar">
<span class="score-label">After</span>
<div class="bar-track"><div class="bar-fill after" style="width:100%"></div></div>
<span class="score-val">100%</span>
</div>
<div class="result-insight">Explicit constraints (char limits, required fields) fix format drift.</div>
</div>
</div>
</div>
</section>
<!-- v2 Features -->
<section>
<div class="container">
<h2>Built to be <span class="highlight">Robust</span></h2>
<p class="section-sub">Production-grade safeguards so you can run it unattended overnight.</p>
<div class="feature-list">
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>Setup Wizard</h4>
<p>6-step wizard with validation gates. Each step must pass before proceeding.</p>
</div>
</div>
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>Dry-Run Gate</h4>
<p>Mandatory pre-flight check. The loop only starts if evals and metrics actually work.</p>
</div>
</div>
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>Train/Test Split</h4>
<p>60/40 split prevents overfitting to specific eval cases.</p>
</div>
</div>
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>Coverage Matrix</h4>
<p>Tracks which improvement categories have been tried, with saturation detection.</p>
</div>
</div>
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>TSV Experiment Log</h4>
<p>One-line-per-experiment flat log. Quick monitoring with tail, grep, awk.</p>
</div>
</div>
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>Crash Recovery</h4>
<p>Classifies errors, retries once, then skips. 3 consecutive crashes pause the loop.</p>
</div>
</div>
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>Exploration-Exploitation</h4>
<p>Early rounds explore broadly, late rounds focus on categories with the best results.</p>
</div>
</div>
<div class="feature">
<div class="feature-dot"></div>
<div>
<h4>Eval Rotation</h4>
<p>Fresh eval queries after 5 experiments to maintain diversity.</p>
</div>
</div>
</div>
</div>
</section>
<!-- Quick Start -->
<section id="quickstart">
<div class="container">
<h2>Quick <span class="highlight">Start</span></h2>
<div class="code-section">
<div class="code-header">
<span class="dot"></span><span class="dot"></span><span class="dot"></span>
<span style="margin-left:8px">Terminal</span>
</div>
<div class="code-body">
<span class="comment"># Clone the repo</span><br>
<span class="cmd">git clone</span> <span class="string">https://github.com/GodModeAI2025/skill-forge.git</span><br><br>
<span class="comment"># Copy into your skills folder (Cowork or OpenClaw)</span><br>
<span class="cmd">cp -r</span> skill-forge/ ~/.skills/skills/skill-forge/<br><br>
<span class="comment"># Auto mode: let it run overnight</span><br>
<span class="string">"Use skill-forge to improve my linkedin-content skill"</span><br><br>
<span class="comment"># Guided mode: decide at every step</span><br>
<span class="string">"Use skill-forge in guided mode to improve my humanizer skill"</span><br><br>
<span class="comment"># Generic mode: optimize any metric</span><br>
<span class="string">"Use skill-forge to optimize train.py — metric: python train.py --eval"</span>
</div>
</div>
</div>
</section>
<!-- Compatibility -->
<section>
<div class="container" style="text-align:center">
<h2>Works with <span class="highlight">your</span> platform</h2>
<p class="section-sub" style="margin:0 auto 32px">Skill Forge uses the standard SKILL.md format and runs wherever AI skills are supported.</p>
<div style="display:flex;justify-content:center;gap:24px;flex-wrap:wrap">
<div style="padding:16px 32px;background:var(--surface);border:1px solid var(--border);border-radius:12px;font-size:15px;font-weight:600">Claude Cowork</div>
<div style="padding:16px 32px;background:var(--surface);border:1px solid var(--border);border-radius:12px;font-size:15px;font-weight:600">OpenClaw</div>
<div style="padding:16px 32px;background:var(--surface);border:1px solid var(--border);border-radius:12px;font-size:15px;font-weight:600;color:var(--muted)">Any SKILL.md host</div>
</div>
</div>
</section>
<!-- Footer -->
<footer>
<div class="container">
<div class="footer-links">
<a href="https://github.com/GodModeAI2025/skill-forge">GitHub</a>
<a href="https://github.com/GodModeAI2025/skill-forge/blob/main/LICENSE">MIT License</a>
<a href="https://github.com/GodModeAI2025/skill-forge/blob/main/SKILL.md">SKILL.md</a>
<a href="https://github.com/GodModeAI2025/skill-forge/blob/main/RELEASE_NOTES.md">Release Notes</a>
</div>
<p>
Built by <a href="https://github.com/GodModeAI2025">Mark Zimmermann</a><br>
Inspired by <a href="https://github.com/karpathy/autoresearch">Karpathy's autoresearch</a><br>
Impressum <a href="https://godmodeai2025.github.io/MarkZimmermann/" target="_blank" rel="noopener">Link</a>
</p>
</div>
</footer>
</body>
</html>