The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
hash	Ready	Preview, Comment	May 11, 2026 3:41pm
hashdotdesign-tokens	Ready	Preview, Comment	May 11, 2026 3:41pm
petrinaut	Ready	Preview, Comment	May 11, 2026 3:41pm

PR Summary

Medium Risk
Medium risk due to behavioral refactors in Petrinaut’s demo app net/handle lifecycle and the simulation timeline’s uPlot update flow, which could affect persistence, undo history, or chart rendering despite being mostly hook/lint-driven changes.

Overview
Adds shared Oxlint React Compiler/React Hooks lint config via .config/oxlint/react-compiler.json, and updates Petrinaut/Refractive/website .oxlintrc.json to extend it (removing per-package react-hooks/* overrides).

Upgrades frontend tooling across the Petrinaut website, @hashintel/petrinaut, and @hashintel/refractive (notably oxlint, oxlint-tsgolint, @typescript/native-preview, vite, plus rolldown/rolldown-plugin-dts and storybook in Petrinaut), with corresponding yarn.lock updates.

Fixes hook/lifecycle issues surfaced by stricter linting: refactors apps/petrinaut-website DevApp to maintain a single active PetrinautDocHandle (instead of per-net cached handles), introduces a default ephemeral net for display, and tightens localStorage syncing/cleanup when switching or creating nets.

Adjusts Petrinaut UI behavior and effects to stabilize dependencies and updates: scroll shadow calculation in VerticalSubViewsContainer, keyboard shortcut listener effect setup, SimulationTimeline uPlot data/update flow (dropping explicit revision prop and updating data based on store length/data changes), adds tabIndex={-1} to list rows for focus management, and uses useEffectEvent in SDCPNView for load-time fitView without refitting on every minZoom change.

^{Reviewed by Cursor Bugbot for commit 6e9bd3a. Bugbot is set up for automated code reviews on this repo. Configure here.}

Merging this PR will not alter performance

✅ 80 untouched benchmarks

_{Comparing cf/fe-627-update-petrinaut-tech-stack-and-dev-tooling (6e9bd3a) with main (f8eae3a)¹}

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 62.29%. Comparing base (f8eae3a) to head (6e9bd3a).
⚠️ Report is 3 commits behind head on main.

@@           Coverage Diff           @@
##             main    #8708   +/-   ##
=======================================
  Coverage   62.29%   62.29%           
=======================================
  Files        1353     1353           
  Lines      136575   136575           
  Branches     5788     5788           
=======================================
+ Hits        85085    85086    +1     
+ Misses      50582    50581    -1     
  Partials      908      908           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

Haven't run/tested the changes locally, but as far as I can tell from a quick skim it all looks good.

If you preferred not to include/write your own oxlint rule .json file for react-compiler though this project has all the eslint-recommended packs transformed for oxlint: https://github.com/popup-plus/oxlint-config-presets

@alex-e-leon Yeah would definitely be better to use existing configs.
I don't have the time to do these changes right now, as I know the choice of the preset will be subject to debate.

If you want to create a PR for this go ahead!

@alex-e-leon Yeah would definitely be better to use existing configs. I don't have the time to do these changes right now, as I know the choice of the preset will be subject to debate.

If you want to create a PR for this go ahead!

For eslint-plugin-react-hooks there's only really one preset option to choose from which is react-hooks/recommended so I think that should prevent any debate : ) (technically there is also react-hooks/recommended-latest but I think we can all agree we don't want to be guinea pigs for rules that are still experiments)

Happy to do it some time later as well though - totally understand if you don't have time to action right now - neither do I! : )

Benchmark results