[sebyx07]

• rake clobber deleted all of tmp/, including the demo app's tracked .keep files, and Fully agent-managed: every CI job required, monthly canary + improve, Dependabot fixes #7 committed that by accident. Restored, and the rake-compiler tmp_dir is now tmp/ext, so clobber only removes build output.
• .claude/settings.json: the ask list prompted on every commit and push. Routine git/gh/rake/bin commands are allowed; git tag, force pushes, rake release, gem push/yank and gh release are denied outright, because merging a version bump publishes via release.yml.

Verified: rake clobber compile keeps the .keep files; CI=1 bundle exec rspec spec/gem passes 225/0; settings JSON valid.

🤖 Generated with Claude Code

[github-actions]

Benchmarks: aarch64 (linux)

ruby 3.4.7 (2025-10-08 revision 7a5688e2a2) +PRISM [aarch64-linux] · commit caa959512a · 2.0s per report
Δ is against main 83f0da849f (same runner type; expect a few % noise).

escape_bench (C ext) · peak RSS 117 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
escape_html short clean (10B)	591,451 ± 0.2%	0.0017	0.00	fastest	-1.1%
escape_html medium clean (200B)	326,600 ± 0.2%	0.0031	0.00	1.81x slower	+1.3%
append short clean (10B)	185,766 ± 0.3%	0.0054	0.02	3.18x slower	+0.2%
append medium clean (200B)	147,841 ± 0.1%	0.0068	0.00	4.00x slower	+1.6%
append short dirty (10B)	144,316 ± 0.3%	0.0069	0.00	4.10x slower	+0.5%
append medium dirty (200B)	74,334 ± 0.1%	0.0135	0.00	7.96x slower	+0.9%
escape_html medium dirty (200B)	55,543 ± 0.4%	0.0180	100	10.65x slower	+1.0%
escape_html short dirty (10B)	48,723 ± 1.5%	0.0205	100	12.14x slower	+10.2%
escape_html long clean (10KB)	11,305 ± 0.1%	0.0885	0.00	52.32x slower	-0.2%
append long clean (10KB)	9,287 ± 0.3%	0.1077	0.00	63.69x slower	+0.1%
append long dirty (10KB)	1,176 ± 0.3%	0.8503	0.00	502.93x slower	+0.6%
escape_html long dirty (10KB)	1,020 ± 7.4%	0.9806	100	579.98x slower	+1.1%

escape_bench (pure Ruby) · peak RSS 150 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
escape_html short clean (10B)	154,047 ± 0.2%	0.0065	0.00	fastest	-0.0%
append short clean (10B)	106,045 ± 0.4%	0.0094	0.02	1.45x slower	-0.5%
escape_html medium clean (200B)	55,530 ± 0.2%	0.0180	0.00	2.77x slower	+0.1%
append medium clean (200B)	47,350 ± 0.2%	0.0211	0.00	3.25x slower	-0.0%
escape_html short dirty (10B)	32,259 ± 0.5%	0.0310	100	4.78x slower	+2.0%
append short dirty (10B)	29,807 ± 0.6%	0.0335	100	5.17x slower	+2.2%
escape_html medium dirty (200B)	23,728 ± 0.2%	0.0421	200	6.49x slower	+3.3%
append medium dirty (200B)	21,781 ± 0.4%	0.0459	200	7.07x slower	+8.4%
escape_html long clean (10KB)	1,629 ± 0.2%	0.6137	0.00	94.54x slower	-0.1%
append long clean (10KB)	1,560 ± 0.3%	0.6409	0.00	98.73x slower	+0.3%
escape_html long dirty (10KB)	1,064 ± 2.3%	0.9401	200	144.82x slower	+0.4%
append long dirty (10KB)	875 ± 9.4%	1.1422	200	175.96x slower	+1.8%

render_bench (C ext) · peak RSS 130 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	7,000 ± 0.4%	0.1429	713	fastest	+13.6%
ruby2html	5,617 ± 0.9%	0.1780	419	1.25x slower	+1.8%
phlex	502 ± 0.6%	1.9939	6,375	13.96x slower	+21.7%

render_bench (pure Ruby) · peak RSS 131 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	7,032 ± 0.6%	0.1422	713	fastest	+14.8%
ruby2html	2,310 ± 0.6%	0.4329	1,029	3.04x slower	+3.4%
phlex	502 ± 1.2%	1.9929	6,375	14.01x slower	+11.0%

tag_bench (C ext) · peak RSS 42 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	222,741 ± 0.5%	0.0045	7.00	fastest	+0.4%
li(text)	150,459 ± 0.4%	0.0066	7.02	1.48x slower	-0.5%
li(dirty text)	122,227 ± 0.4%	0.0082	7.00	1.82x slower	+0.1%
div { block }	108,364 ± 0.5%	0.0092	7.00	2.06x slower	-0.1%
span(text, class:)	29,288 ± 1.1%	0.0341	207	7.61x slower	+2.1%
div(class:) { block }	27,214 ± 0.4%	0.0367	207	8.18x slower	+2.2%
a(text, href:, class:)	22,380 ± 0.3%	0.0447	207	9.95x slower	+0.6%

tag_bench (pure Ruby) · peak RSS 26 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	81,843 ± 0.4%	0.0122	107	fastest	+3.8%
div { block }	62,998 ± 0.4%	0.0159	107	1.30x slower	+9.5%
li(text)	54,170 ± 0.4%	0.0185	107	1.51x slower	+1.6%
li(dirty text)	40,603 ± 0.5%	0.0246	207	2.02x slower	+3.8%
div(class:) { block }	14,753 ± 0.7%	0.0678	207	5.55x slower	+2.0%
span(text, class:)	14,558 ± 0.3%	0.0687	207	5.62x slower	+2.1%
a(text, href:, class:)	9,424 ± 0.5%	0.1061	207	8.68x slower	+1.9%

Rails request (C ext) · peak RSS 221 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
Ruby2html::Component	418 ± 1.2%	2.3914	4,703	fastest	-1.4%
ruby2html .html.rb templates	412 ± 1.7%	2.4278	4,846	1.02x slower	-2.8%
Slim	408 ± 4.4%	2.4535	4,375	1.03x slower	-1.5%
ViewComponent + ruby2html	399 ± 1.3%	2.5032	4,561	1.05x slower	+0.4%
ERB	395 ± 4.3%	2.5305	3,130	1.06x slower	+3.5%
Phlex	375 ± 1.3%	2.6673	6,127	1.12x slower	-2.7%

[github-actions]

Benchmarks: x86_64 (linux)

ruby 3.4.7 (2025-10-08 revision 7a5688e2a2) +PRISM [x86_64-linux] · commit caa959512a · 2.0s per report
Δ is against main 83f0da849f (same runner type; expect a few % noise).

escape_bench (C ext) · peak RSS 118 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
escape_html short clean (10B)	595,222 ± 0.5%	0.0017	0.00	fastest	+0.5%
escape_html medium clean (200B)	378,084 ± 0.5%	0.0026	0.00	1.57x slower	+0.4%
append short clean (10B)	213,863 ± 0.5%	0.0047	0.02	2.78x slower	+1.2%
append medium clean (200B)	162,065 ± 3.1%	0.0062	0.00	3.67x slower	-1.1%
append short dirty (10B)	149,049 ± 1.8%	0.0067	0.00	3.99x slower	-0.1%
append medium dirty (200B)	94,388 ± 0.5%	0.0106	0.00	6.31x slower	-0.4%
escape_html medium dirty (200B)	62,485 ± 7.5%	0.0160	100	9.53x slower	-4.4%
escape_html short dirty (10B)	44,313 ± 1.2%	0.0226	100	13.43x slower	-2.3%
escape_html long clean (10KB)	16,072 ± 1.3%	0.0622	0.00	37.03x slower	-0.4%
append long clean (10KB)	10,777 ± 2.8%	0.0928	0.00	55.23x slower	-6.1%
append long dirty (10KB)	1,496 ± 0.3%	0.6685	0.00	397.89x slower	-0.3%
escape_html long dirty (10KB)	1,256 ± 11.0%	0.7962	100	473.89x slower	+0.4%

escape_bench (pure Ruby) · peak RSS 154 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
escape_html short clean (10B)	210,421 ± 0.4%	0.0048	0.00	fastest	+0.5%
append short clean (10B)	145,061 ± 0.7%	0.0069	0.02	1.45x slower	+1.4%
escape_html medium clean (200B)	87,136 ± 0.9%	0.0115	0.00	2.41x slower	+0.7%
append medium clean (200B)	72,754 ± 1.9%	0.0137	0.00	2.89x slower	+2.1%
escape_html short dirty (10B)	33,616 ± 1.3%	0.0297	100	6.26x slower	-0.6%
append short dirty (10B)	30,400 ± 2.1%	0.0329	100	6.92x slower	-0.7%
escape_html medium dirty (200B)	25,421 ± 0.8%	0.0393	200	8.28x slower	-5.2%
append medium dirty (200B)	23,303 ± 1.1%	0.0429	200	9.03x slower	-4.1%
escape_html long clean (10KB)	3,069 ± 0.5%	0.3259	0.00	68.57x slower	+0.1%
append long clean (10KB)	2,786 ± 0.6%	0.3590	0.00	75.54x slower	-0.3%
escape_html long dirty (10KB)	1,313 ± 2.6%	0.7619	200	160.32x slower	-2.4%
append long dirty (10KB)	1,007 ± 11.9%	0.9930	200	208.95x slower	-6.9%

render_bench (C ext) · peak RSS 131 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	6,636 ± 2.4%	0.1507	713	fastest	-0.9%
ruby2html	5,142 ± 1.0%	0.1945	419	1.29x slower	-1.4%
phlex	428 ± 1.9%	2.3362	6,375	15.50x slower	-6.3%

render_bench (pure Ruby) · peak RSS 131 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	4,815 ± 2.7%	0.2077	713	fastest	-3.0%
ruby2html	1,801 ± 2.5%	0.5553	1,029	2.67x slower	-6.0%
phlex	419 ± 1.0%	2.3863	6,375	11.49x slower	-1.4%

tag_bench (C ext) · peak RSS 43 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	238,878 ± 1.0%	0.0042	7.00	fastest	-0.9%
li(text)	155,910 ± 3.0%	0.0064	7.02	1.53x slower	-2.0%
li(dirty text)	124,790 ± 2.1%	0.0080	7.00	1.91x slower	+6.4%
div { block }	121,827 ± 2.0%	0.0082	7.00	1.96x slower	-1.1%
span(text, class:)	26,742 ± 0.6%	0.0374	207	8.93x slower	-0.9%
div(class:) { block }	25,415 ± 1.6%	0.0393	207	9.40x slower	-1.0%
a(text, href:, class:)	19,936 ± 6.3%	0.0502	207	11.98x slower	-4.2%

tag_bench (pure Ruby) · peak RSS 27 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	84,050 ± 1.0%	0.0119	107	fastest	-0.7%
div { block }	67,647 ± 1.3%	0.0148	107	1.24x slower	-5.0%
li(text)	64,026 ± 1.3%	0.0156	107	1.31x slower	-1.4%
li(dirty text)	44,849 ± 1.7%	0.0223	207	1.87x slower	-2.2%
span(text, class:)	13,939 ± 0.7%	0.0717	207	6.03x slower	-1.3%
div(class:) { block }	13,372 ± 0.8%	0.0748	207	6.29x slower	-1.5%
a(text, href:, class:)	9,462 ± 0.6%	0.1057	207	8.88x slower	-3.2%

Rails request (C ext) · peak RSS 213 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
Slim	522 ± 7.9%	1.9156	3,953	fastest	-3.3%
Ruby2html::Component	502 ± 2.0%	1.9912	4,267	1.04x slower	-2.9%
ruby2html .html.rb templates	499 ± 1.8%	2.0054	4,403	1.05x slower	-2.1%
ViewComponent + ruby2html	479 ± 1.7%	2.0888	4,125	1.09x slower	-3.1%
Phlex	461 ± 2.0%	2.1706	5,537	1.13x slower	-2.2%
ERB	461 ± 11.9%	2.1715	2,854	1.13x slower	-4.8%