[jdx]

🤖 Refreshed benchmarks

benchmarks/results.json was pinned to aube 1.26.0; the workspace is now 1.27.0. Re-ran mise run bench:bump on the hermetic Verdaccio registry (500mbit / 50ms per the mise task) and regenerated benchmarks/results.json plus the README BENCH_RATIOS block. The benchmark matrix pins aube's GVS mode via npm_config_enable_global_virtual_store=true|false (the auto-synthesized env alias for the enableGlobalVirtualStore setting), so GitHub Actions' inherited CI=true environment does not change whether aube runs with GVS enabled or disabled.

Benchmark changes

Versions:

• aube: 1.26.0 -> 1.27.0
• deno: 2.9.1 -> 2.9.2
• pnpm: 11.10.0 -> 11.13.0
• yarn: 4.17.0 -> 4.17.1

Public ratios: warm installs vs Bun 3x -> 11x; warm installs vs pnpm 6x -> 13x; repeat test vs Bun 3x -> 6x; repeat test vs pnpm 28x -> 43x.

Benchmark	aube	bun	deno	pnpm	npm	yarn
Fresh install (warm cache)	380ms -> 263ms (-31%)	1154ms -> 2924ms (+153%)	1353ms -> 1292ms (-5%)	2212ms -> 3419ms (+55%)	8656ms -> 7825ms (-10%)	8864ms -> 9920ms (+12%)
Fresh install (cold cache)	17450ms -> 17149ms (-2%)	7541ms -> 6311ms (-16%)	8449ms -> 7838ms (-7%)	14387ms -> 16838ms (+17%)	11424ms -> 11588ms (+1%)	15374ms -> 14444ms (-6%)
npm install && npm run test	15ms -> 10ms (-33%)	48ms -> 55ms (+15%)	71ms -> 76ms (+7%)	419ms -> 432ms (+3%)	849ms -> 853ms (+0%)	1834ms -> 1965ms (+7%)

Review the numbers before merging — if anything looks wildly off vs. the previous release, investigate before landing. Hermetic proxy jitter or an npmjs uplink hiccup can occasionally skew results.

Once merged to main, the updated bench results flow into the next release-plz-pr run automatically.

---

Generated by the bench-refresh workflow.

---

Note

Low Risk
Documentation and benchmark artifact updates only; no runtime or install logic changes.

Overview
Re-runs the hermetic benchmark suite and syncs published numbers with aube 1.27.0.

benchmarks/results.json is replaced with a new run (updated timestamp and pinned tool versions: aube 1.27.0, deno 2.9.2, pnpm 11.13.0, yarn 4.17.1). Timings shift across all three scenarios (warm/cold fresh install, repeat install && test); aube warm install improves (~380ms → ~263ms) while several competitors move in the opposite direction on warm install (notably Bun ~1.2s → ~2.9s), which drives higher relative ratios in marketing copy.

The README.md BENCH_RATIOS block is regenerated to match: warm install claims move from ~6x/~3x vs pnpm/Bun to ~13x/~11x; repeat-test claims move from ~28x/~3x to ~43x/~6x.

^{Reviewed by Cursor Bugbot for commit ab490e8. Bugbot is set up for automated code reviews on this repo. Configure here.}

Summary by CodeRabbit

• Documentation

  • Updated the “Fast installs” benchmark figures in the README to reflect the latest results.

• Chores

  • Refreshed benchmark measurements and tool version metadata for warm, cold, and test-install scenarios.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Central YAML (base), Organization UI (inherited)

Review profile: CHILL

Plan: Pro Plus

Run ID: 79a2deb0-b941-49cc-b7b1-e64085da4d52

📥 Commits

Reviewing files that changed from the base of the PR and between d881e88 and ab490e8.

📒 Files selected for processing (2)

• README.md
• benchmarks/results.json

---

📝 Walkthrough

Walkthrough

Benchmark metadata, tool versions, measurement outputs, and the README’s Fast installs performance figures were refreshed.

Changes

Benchmark refresh

Layer / File(s)	Summary
Refresh benchmark measurements benchmarks/results.json	Updated the benchmark timestamp and tool versions, then replaced the gvs-warm, gvs-cold, and install-test values and statistics.
Update README benchmark claims README.md	Updated the Fast installs benchmark multipliers with the refreshed figures.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related PRs

• jdx/aube#1002: Refreshes the same benchmark results and README ratio claims.
• jdx/aube#975: Updates the same benchmark scenarios, versions, and README figures.
• jdx/aube#962: Regenerates benchmark data and corresponding README performance claims.

Suggested reviewers: mise-en-dev

Poem

I’m a bunny with numbers,
Hopping through fresh results bright,
Warm installs leap higher,
Cold runs measure right.
The README follows—
Benchmark carrots in sight!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly matches the main change: refreshing benchmark data for v1.27.0.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[greptile-apps]

Greptile Summary

This PR refreshes benchmarks/results.json and the README.md BENCH_RATIOS block for aube v1.27.0, bumping companion tool versions (deno 2.9.2, pnpm 11.13.0, yarn 4.17.1) and regenerating all timing data from a hermetic Verdaccio run.

• The JSON values are internally consistent — every values entry matches its corresponding stats.mean when rounded to milliseconds, and all ratio claims in the README are arithmetically correct against the new data.
• Bun's warm-cache time rose from 1154 ms to 2924 ms (+153 %), which is the primary driver of the headline changing from "3x vs Bun" to "11x vs Bun"; this looks more like a proxy/network hiccup than a real Bun regression and may overstate aube's advantage in the README.
• The pnpm cold-install stddev jumped from 0.411 s to 5.149 s (12.5x), with a min/max spread of ~10 s, suggesting at least one cold run was contaminated by upstream jitter.

Confidence Score: 4/5

Safe to merge if the benchmark numbers have been manually reviewed; the only risk is that noisy measurements flow into public-facing marketing copy.

The change is purely data — no executable logic is modified. All internal consistency checks pass (values match means, README ratios match JSON). The two concerns are about measurement reliability: Bun's warm-cache result looks anomalously slow and pnpm's cold-install variance is very high. If those numbers are accepted as-is, the README will advertise ratios that may be inflated by a bad benchmark run rather than reflecting typical performance.

benchmarks/results.json — specifically the gvs-warm Bun entry and the gvs-cold pnpm entry, which show outlier behavior worth re-running before the ratios are published in the README.

Important Files Changed

Filename	Overview
benchmarks/results.json	Benchmark data updated to aube 1.27.0. Internal value/mean consistency is correct, but Bun warm-cache shows a suspicious +153% regression and pnpm cold-install stddev is 12.5x higher than before, both of which may be measurement artifacts worth investigating before the README claims flow downstream.
README.md	BENCH_RATIOS block updated; ratios are arithmetically correct against the new JSON values, but the "11x vs Bun" claim derives from an anomalous Bun warm-cache result that may overstate the advantage.

_{Reviews (1): Last reviewed commit: "chore: refresh benchmarks" | Re-trigger Greptile}

[greptile-apps]

Bun warm-cache result looks anomalous

Bun's warm-cache time jumped from 1154 ms to 2924 ms (+153 %), which is the primary driver of the headline claim changing from "3x faster than Bun" to "11x faster than Bun". A 153 % regression in a single Bun release seems unlikely to be a real improvement; it's more consistent with a proxy/network hiccup during the Bun run. The stddev for this run (261 ms over a ~2.9 s mean) is tighter than before, which rules out per-sample noise — the whole batch simply ran slow. Retrying just the gvs-warm row for Bun before merging would confirm whether this is real or a measurement artifact.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

[greptile-apps]

pnpm cold-install stddev is unusually high

The pnpm cold-install stddev jumped from 0.411 s to 5.149 s (12.5×), with a min/max spread of 12.48 s–22.74 s. That 10-second range on a ~16.8 s mean suggests at least one run was seriously impacted by an upstream hiccup rather than a clean cold-cache scenario. The mean itself shifted up (+17 %), but with that variance the measurement is not reliable enough to draw conclusions about pnpm's cold performance. A re-run of the gvs-cold row would tighten the stddev and either confirm or discount the regression.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!