[developerz-ai]

What

mise.toml duplicated .ruby-version's literal (ruby = "4.0.9"), so the two could
drift silently: bin/setup's mise install would install the stale mise.toml
version while CI ran the bumped .ruby-version. mise.toml now reads the pin
from .ruby-version via a read_file template, bin/setup stays a bare 'mise
install' so the installed toolchain resolves through mise.toml from the same
single source, and a new spec/bin/ruby_version_spec.rb fails when any of the
three files reintroduces an independent version literal.

Acceptance

• mise.toml states no Ruby version literal: its [tools] ruby entry is the read_file template of .ruby-version, so editing .ruby-version alone changes the version mise resolves and installs.
• bin/setup keeps its single bare system!('mise install') call and gains no version literal; the toolchain it installs resolves mise.toml → .ruby-version.
• spec/bin/ruby_version_spec.rb passes on the changed tree and fails if mise.toml re-adds a literal or references another file, if .ruby-version stops being an exact X.Y.Z, or if bin/setup passes an explicit ruby@ version to mise.
• CI=1 bundle exec rspec includes and passes the new spec, and bundle exec rubocop is clean on the touched Ruby file.

Changes

• mise.toml (edit) — Replace the duplicated literal on line 4 (ruby = "4.0.9") with ruby = "{{ read_file(path='.ruby-version') | trim }}" and reword the two header comments to state that .ruby-version is the single source of the Ruby pin and that bin/setup's mise install resolves it. Keep the [tools] table and everything else unchanged.
• spec/bin/ruby_version_spec.rb (create) — New regression spec enforcing the single-source Ruby pin contract from sharedContext: (1) .ruby-version is an exact X.Y.Z after strip; (2) the ruby request in mise.toml must reference .ruby-version via read_file(path='.ruby-version') — a version literal or any other referenced file fails — and the referenced file's stripped content must equal .ruby-version; (3) bin/setup's only system!('mise ...') call is exactly 'mise install' with no version argument. Style per spec/bin/setup_spec.rb.

Verification

./bin/check passed (exit 0) in 24s.

What the editors ran to see the change working, and what they saw:

• code_changes paths=["mise.toml"] (toolchain): modified; diff shows ruby = "4.0.9" replaced by ruby = "{{ read_file(path='.ruby-version') | trim }}", header comments reworded, [tools] table unchanged; no runtime check run (brief scoped none)
• bundle exec rspec spec/bin/ruby_version_spec.rb (tests): 4 examples, 0 failures
• drift matrix: same spec run against mutated scratch copies in /tmp (literal in mise.toml; other referenced file; .ruby-version=4.0; system!('mise install ruby@4.0.9')) (tests): control PASS; all four drift scenarios FAIL with the corresponding expectation failing (e.g. expected "4.0.9" to include "read_file(path='.ruby-version')", expected "4.0" to match /\A\d+.\d+.\d+\z/)

Notes

Premises of the request the team disproved:

• The request implies bin/setup itself must gain version-resolution logic; survey shows bin/setup:27 already runs bare 'mise install' and resolves transitively through mise.toml, so once mise.toml derives from .ruby-version bin/setup needs no edit and must NOT gain a second resolver (declared, listed in noChangesNeeded).
• Mise cannot be pointed at .ruby-version via a 'file:' request prefix or via idiomatic version files without per-machine settings: mise docs confirm no file: prefix exists and idiomatic_version_file_enable_tools is empty by default, so the only project-local derivation is the read_file template used here.

---

🤖 Opened by developerz.ai for task tsk_640893e36471b2e8dffe60bdbb0599b6 on sebyx07/ruby2html.

Raised automatically by the raw_task_triage lane.

🤖 developerz.ai, an automated agent, not a human — this PR is one slice of a stacked plan.
Dispatched automatically by a developerz.ai maintainer lane
Part 1/1

[github-actions]

Benchmarks: x86_64 (linux)

ruby 4.0.9 (2026-10-23 revision 3e93b23afc) +PRISM [x86_64-linux] · commit 23ca711a8d · 2.0s per report
Δ is against main 09f01f322e (same runner type; expect a few % noise).

escape_bench (C ext) · peak RSS 119 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
escape_html short clean (10B)	704,872 ± 0.6%	0.0014	0.00	fastest	-20.2%
escape_html medium clean (200B)	418,690 ± 0.4%	0.0024	0.00	1.68x slower	-21.4%
append short clean (10B)	255,479 ± 2.6%	0.0039	0.02	2.76x slower	-23.9%
append short dirty (10B)	186,529 ± 1.0%	0.0054	0.00	3.78x slower	-24.6%
append medium clean (200B)	167,573 ± 3.8%	0.0060	0.00	4.21x slower	-27.2%
escape_html short dirty (10B)	128,196 ± 1.8%	0.0078	100	5.50x slower	-22.6%
append medium dirty (200B)	91,728 ± 0.7%	0.0109	0.00	7.68x slower	-22.2%
escape_html medium dirty (200B)	71,789 ± 1.1%	0.0139	100	9.82x slower	-21.9%
escape_html long clean (10KB)	16,228 ± 0.2%	0.0616	0.00	43.43x slower	-12.5%
append long clean (10KB)	11,476 ± 0.4%	0.0871	0.00	61.42x slower	-19.6%
append long dirty (10KB)	1,515 ± 0.3%	0.6603	0.00	465.41x slower	-19.6%
escape_html long dirty (10KB)	1,459 ± 3.0%	0.6856	100	483.26x slower	-20.7%

escape_bench (pure Ruby) · peak RSS 145 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
escape_html short clean (10B)	200,792 ± 0.5%	0.0050	0.00	fastest	-24.4%
append short clean (10B)	151,218 ± 2.6%	0.0066	0.02	1.33x slower	-24.0%
escape_html short dirty (10B)	84,528 ± 1.7%	0.0118	100	2.38x slower	-28.7%
append short dirty (10B)	64,014 ± 3.3%	0.0156	100	3.14x slower	-35.5%
escape_html medium clean (200B)	57,454 ± 0.3%	0.0174	0.00	3.49x slower	-17.9%
append medium clean (200B)	48,554 ± 2.9%	0.0206	0.00	4.14x slower	-20.6%
escape_html medium dirty (200B)	25,368 ± 1.9%	0.0394	200	7.92x slower	-34.3%
append medium dirty (200B)	22,116 ± 1.3%	0.0452	200	9.08x slower	-36.0%
escape_html long clean (10KB)	1,552 ± 0.3%	0.6442	0.00	129.35x slower	-11.9%
append long clean (10KB)	1,484 ± 0.4%	0.6737	0.00	135.27x slower	-11.5%
escape_html long dirty (10KB)	1,014 ± 1.7%	0.9864	200	198.06x slower	-15.9%
append long dirty (10KB)	902 ± 3.2%	1.1086	200	222.59x slower	-18.8%

render_bench (C ext) · peak RSS 132 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	8,065 ± 2.6%	0.1240	713	fastest	-28.5%
ruby2html	5,414 ± 1.1%	0.1847	419	1.49x slower	-33.6%
phlex	498 ± 2.2%	2.0090	6,374	16.20x slower	-30.9%

render_bench (pure Ruby) · peak RSS 132 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	7,985 ± 1.6%	0.1252	713	fastest	-28.8%
ruby2html	2,199 ± 0.7%	0.4547	1,029	3.63x slower	-30.6%
phlex	503 ± 2.4%	1.9882	6,374	15.88x slower	-29.8%

tag_bench (C ext) · peak RSS 40 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	251,108 ± 0.8%	0.0040	7.00	fastest	-29.1%
li(text)	170,568 ± 1.9%	0.0059	7.02	1.47x slower	-23.4%
li(dirty text)	144,584 ± 0.5%	0.0069	7.00	1.74x slower	-25.0%
div { block }	112,539 ± 1.1%	0.0089	7.00	2.23x slower	-25.8%
span(text, class:)	31,573 ± 1.7%	0.0317	207	7.95x slower	-34.3%
div(class:) { block }	27,147 ± 1.2%	0.0368	207	9.25x slower	-37.3%
a(text, href:, class:)	22,572 ± 1.5%	0.0443	207	11.12x slower	-35.1%

tag_bench (pure Ruby) · peak RSS 27 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	97,501 ± 0.8%	0.0103	107	fastest	-28.2%
div { block }	79,460 ± 0.8%	0.0126	107	1.23x slower	-27.7%
li(text)	71,926 ± 1.1%	0.0139	107	1.36x slower	-25.7%
li(dirty text)	47,789 ± 4.2%	0.0209	207	2.04x slower	-30.4%
span(text, class:)	14,321 ± 0.7%	0.0698	207	6.81x slower	-31.9%
div(class:) { block }	13,461 ± 2.7%	0.0743	207	7.24x slower	-31.3%
a(text, href:, class:)	9,532 ± 0.6%	0.1049	207	10.23x slower	-31.5%

Rails request (C ext) · peak RSS 242 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
Ruby2html::Component	509 ± 3.3%	1.9649	4,347	fastest	-29.5%
ruby2html .html.rb templates	504 ± 3.6%	1.9822	4,474	1.01x slower	-30.6%
Slim	500 ± 8.8%	2.0008	4,039	1.02x slower	-34.7%
ViewComponent + ruby2html	478 ± 3.1%	2.0918	4,204	1.06x slower	-31.2%
ERB	435 ± 17.7%	2.2997	2,887	1.17x slower	-36.3%
Phlex	435 ± 3.7%	2.3011	5,648	1.17x slower	-32.2%

[github-actions]

Benchmarks: aarch64 (linux)

ruby 4.0.9 (2026-10-23 revision 3e93b23afc) +PRISM [aarch64-linux] · commit 23ca711a8d · 2.0s per report
Δ is against main 09f01f322e (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)	610,588 ± 0.2%	0.0016	0.00	fastest	-1.2%
escape_html medium clean (200B)	336,839 ± 0.2%	0.0030	0.00	1.81x slower	+0.0%
append short clean (10B)	208,668 ± 0.3%	0.0048	0.02	2.93x slower	-0.6%
append short dirty (10B)	160,861 ± 0.2%	0.0062	0.00	3.80x slower	-0.0%
append medium clean (200B)	147,217 ± 0.2%	0.0068	0.00	4.15x slower	-0.5%
escape_html short dirty (10B)	109,545 ± 1.3%	0.0091	100	5.57x slower	-1.2%
append medium dirty (200B)	73,674 ± 0.2%	0.0136	0.00	8.29x slower	-0.1%
escape_html medium dirty (200B)	58,331 ± 0.4%	0.0171	100	10.47x slower	-0.6%
escape_html long clean (10KB)	11,460 ± 0.2%	0.0873	0.00	53.28x slower	+0.0%
append long clean (10KB)	8,958 ± 0.3%	0.1116	0.00	68.16x slower	-1.0%
append long dirty (10KB)	1,202 ± 0.3%	0.8317	0.00	507.82x slower	-0.0%
escape_html long dirty (10KB)	1,125 ± 3.3%	0.8890	100	542.83x slower	-1.4%

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)	164,582 ± 0.2%	0.0061	0.00	fastest	-0.0%
append short clean (10B)	104,426 ± 0.2%	0.0096	0.02	1.58x slower	-0.0%
escape_html short dirty (10B)	69,249 ± 1.0%	0.0144	100	2.38x slower	-1.6%
append short dirty (10B)	57,660 ± 0.6%	0.0173	100	2.85x slower	-1.0%
escape_html medium clean (200B)	56,845 ± 0.2%	0.0176	0.00	2.90x slower	-0.1%
append medium clean (200B)	45,650 ± 0.2%	0.0219	0.00	3.61x slower	+0.3%
escape_html medium dirty (200B)	23,965 ± 2.6%	0.0417	200	6.87x slower	-5.5%
append medium dirty (200B)	21,767 ± 0.4%	0.0459	200	7.56x slower	-4.1%
escape_html long clean (10KB)	1,632 ± 0.2%	0.6126	0.00	100.83x slower	-0.1%
append long clean (10KB)	1,560 ± 0.3%	0.6412	0.00	105.53x slower	-0.2%
escape_html long dirty (10KB)	1,084 ± 0.6%	0.9224	200	151.81x slower	-1.0%
append long dirty (10KB)	964 ± 2.4%	1.0375	200	170.75x slower	-1.7%

render_bench (C ext) · peak RSS 131 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	7,592 ± 0.7%	0.1317	713	fastest	-2.3%
ruby2html	5,800 ± 0.6%	0.1724	419	1.31x slower	-1.2%
phlex	565 ± 0.9%	1.7694	6,374	13.43x slower	-3.1%

render_bench (pure Ruby) · peak RSS 131 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
erubi	7,499 ± 1.3%	0.1334	713	fastest	-2.7%
ruby2html	2,336 ± 0.7%	0.4282	1,029	3.21x slower	-2.0%
phlex	563 ± 0.7%	1.7752	6,374	13.31x slower	-4.3%

tag_bench (C ext) · peak RSS 39 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	218,694 ± 0.7%	0.0046	7.00	fastest	-0.4%
li(text)	157,151 ± 1.0%	0.0064	7.02	1.39x slower	-2.8%
li(dirty text)	124,918 ± 0.4%	0.0080	7.00	1.75x slower	-1.5%
div { block }	98,874 ± 0.7%	0.0101	7.00	2.21x slower	-0.9%
span(text, class:)	31,115 ± 0.4%	0.0321	207	7.03x slower	-0.1%
div(class:) { block }	29,659 ± 0.3%	0.0337	207	7.37x slower	+0.3%
a(text, href:, class:)	23,891 ± 0.3%	0.0419	207	9.15x slower	+0.1%

tag_bench (pure Ruby) · peak RSS 26 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
br	81,141 ± 1.5%	0.0123	107	fastest	+0.2%
div { block }	62,970 ± 0.4%	0.0159	107	1.29x slower	-0.1%
li(text)	54,208 ± 1.4%	0.0184	107	1.50x slower	+0.7%
li(dirty text)	41,332 ± 0.6%	0.0242	207	1.96x slower	+0.8%
div(class:) { block }	15,732 ± 0.4%	0.0636	207	5.16x slower	+0.7%
span(text, class:)	15,482 ± 1.9%	0.0646	207	5.24x slower	-0.2%
a(text, href:, class:)	9,853 ± 0.4%	0.1015	207	8.24x slower	+0.7%

Rails request (C ext) · peak RSS 228 MB

Benchmark	i/s	ms / call	allocs / call	vs fastest	Δ vs main
Ruby2html::Component	465 ± 1.1%	2.1522	4,470	fastest	+1.1%
ruby2html .html.rb templates	464 ± 1.3%	2.1565	4,612	1.00x slower	+1.6%
Slim	459 ± 4.4%	2.1802	4,132	1.01x slower	+1.6%
ViewComponent + ruby2html	435 ± 1.2%	2.3013	4,327	1.07x slower	+0.7%
ERB	425 ± 9.2%	2.3507	2,946	1.09x slower	-1.3%
Phlex	404 ± 1.5%	2.4733	5,839	1.15x slower	+0.5%

[developerz-ai]

✅ Reviewed — nothing blocking

0 actionable comment(s) · grounded on your code

_{⏱ 45s wall clock · glm-5.3-flash via router · 2 model call(s) · 575 output token(s)}

_{🤖 developerz.ai — automated review, running on your model and your box.}

[developerz-ai]

Review summary — 2 file(s), 0 finding(s).

Replaces mise.toml's duplicated Ruby literal with a read_file template of .ruby-version and adds a regression spec guarding the single-source pin; the spec's regexes correctly fail on literals, other referenced files, non-X.Y.Z pins and versioned mise args.

• concern-tests: nothing to review — The added spec asserts the contract directly (X.Y.Z shape, read_file template referencing .ruby-version, resolved content equality, bare 'mise install') and the mise.toml template matches it; no loosened assertions or untested behaviour in the diff.

🤖 developerz.ai review — automated, running on your model and your box. What is this? · Reviewed by router/glm-5.3-flash — 22.8k in / 575 out over 2 calls, on your flat-rate plan: no per-token charge (~<$0.01 at list price).