[ci] Benchmark tracking

[clabby]

Overview

Adds a new workflow for running a configurable list of benchmarks on PRs and comparing them with results from main in order for us to more easily catch performance regressions in important areas.

[cloudflare-workers-and-pages]

Deploying with    Cloudflare Workers

The latest updates on your project. Learn more about integrating Git with Workers.

Status	Name	Latest Commit	Updated (UTC)
✅ Deployment successful! View logs	commonware-mcp	ab6116b	May 04 2026, 06:43 PM

[cloudflare-workers-and-pages]

Deploying monorepo with    Cloudflare Pages

Latest commit:	ab6116b
Status:	✅  Deploy successful!
Preview URL:	https://a9053fbe.monorepo-eu0.pages.dev
Branch Preview URL:	https://cl-bench-tracker.monorepo-eu0.pages.dev

View logs

[github-actions]

Benchmark results

Tip

✅ PASSED: No benchmark exceeded the regression threshold.

Warning

1 benchmark(s) had no uploaded main-branch baseline.

Benchmark comparison table

Benchmark	Baseline (main)	Current	Delta	Threshold	Status
qmdb::merkleize/variant=any::unordered::fixed::mmr keys=10000 ch=false sync=false	-	1.533 ms	-	10.00%	⚠️ WARN missing baseline

[patrick-ogrady]

As discussed, we should probably resolve this:

- [P2] Gate PR comments to write-capable runs — /Users/patrickogrady/code/monorepo-pr-3698/.github/workflows/benchmark-tracking.yml:101-101
    For pull_request runs from forks or Dependabot, GitHub downgrades GITHUB_TOKEN to read-only even though the workflow requests issues: write, so this comment step will 403 after a successful benchmark run and mark the check
    failed even when there is no benchmark regression. Gate this step to same-repository PRs or make comment posting non-blocking so external PRs can still use the step summary and regression failure logic.

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit ab6116b. Configure here.}

[cursor]

Unconditional cancel-in-progress breaks main baseline uploads

Medium Severity

cancel-in-progress: true is unconditional, unlike every other workflow in this repo which uses cancel-in-progress: ${{ github.event_name == 'pull_request' }}. This means rapid pushes to main will cancel in-progress benchmark runs before they can upload their baseline artifact, potentially leaving PRs with a stale or missing baseline. For a 180-minute-timeout benchmark job, this window is significant.

 

^{Reviewed by Cursor Bugbot for commit ab6116b. Configure here.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 95.77%. Comparing base (52f31c8) to head (ab6116b).
⚠️ Report is 3 commits behind head on main.

@@           Coverage Diff           @@
##             main    #3698   +/-   ##
=======================================
  Coverage   95.77%   95.77%           
=======================================
  Files         459      458    -1     
  Lines      178700   178691    -9     
  Branches     4158     4158           
=======================================
- Hits       171143   171139    -4     
+ Misses       6232     6226    -6     
- Partials     1325     1326    +1     

see 10 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 52f31c8...ab6116b. Read the comment docs.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.