[SandroMaglione]

Summary

Add Oxlint to the repository and run the repository's own Effect Machine plugin on its machines.

Formatter check. Effect-TS/effect main (pushed 2026-10-31) formats with dprint (^0.63.0) and lints with oxlint (^1.88.0): "lint": "oxlint -f unix && dprint check", with dprint.json and .oxlintrc.json at the root and no .oxfmtrc*, biome, or prettier config. Effect-TS/effect-smol and the local beta.98 mirror do the same with dprint 0.55 and oxlint 1.42. So formatting stays on dprint. This PR does not touch the dprint version; the dependency PR bumps it.

Lint setup

• .oxlintrc.json, modeled on the ritmo config: the correctness category plus type-aware rules through oxlint-tsgolint (no-unnecessary-type-assertion, switch-exhaustiveness-check, no-floating-promises, no-misused-promises) and no-unused-vars with _ ignore patterns. Pins oxlint 1.90.0 to match the dependency-bump PR, and oxlint-tsgolint 7.0.2005.
• pnpm lint builds @typeonce/oxlint-plugin-effect-machine (workspace:*, loaded as a jsPlugin) and then runs oxlint, so it works in a fresh clone. It runs in pnpm check right after format:check, so the CI check and latest-effect jobs both run it. I did not add a separate ci.yml step because it would run the same lint twice.
• The plugin's current recommended rules apply to the devtools example machines (packages/devtools/src/internal/browser/{example-machine,*-example}.ts) and to packages/effect-machine/test/{fixtures,examples}. Docs snippets live in Markdown, which Oxlint cannot lint.
• The root tsconfig.json now references tsconfig.tests.json. Without it, tsgolint (and editors) type-checked tests in an inferred project without noUncheckedIndexedAccess, which produced about 180 false "unnecessary !" reports. tsc -b and every script that reads a tsconfig are unaffected.

Rules left out or turned off, and why

• Type tests, type-performance fixtures, and scripts/** belong to no tsconfig, so type-aware rules are off there. Type tests and perf fixtures also skip no-unused-vars: type tests bind values only to check their types, and the perf fixtures stay byte-stable so type-performance comparisons stay valid.
• no-unused-vars is off for packages/*/src/**. Public and internal types keep phantom type parameters for variance and inference, and noUnusedLocals already checks the rest.
• no-unnecessary-type-arguments (in ritmo's set) is left out: Effect-style code spells out Effect.Effect<A, E, never> on purpose.
• no-redundant-type-constituents is off: unions such as unknown | StoppedError and string | "dynamic" document the intended values.
• unicorn/no-new-array is off, as in Effect's config: the execution plan preallocates arrays.
• no-this-alias allows self, the capture idiom for Effect.gen bodies.

Fixes (no behavior change)

• Removed about 100 unnecessary type assertions and non-null assertions in core, devtools, React, and tests. Casts only; the emitted JavaScript stays the same apart from the items below.
• Five assertions stay with a reasoned oxlint-disable-next-line. TypeScript 6.0.3 needs them, but the TypeScript 7 checker in tsgolint does not:
  • the TargetSelectionTypeId key,
  • two places in finiteModel.ts,
  • the ELK CommonJS constructor,
  • one testing-library query.
    The as never in Paths.test.ts also stays: TypeScript 6 rejects the reserved root state name there, but TypeScript 7 accepts it. That may be a type-level gap under TS 7 worth checking in the TS 7 lane.
• Other fixes:
  • ({}) => to () =>, and { input: input } to { input }.
  • Removed ...{ literal } and array spreads that did nothing.
  • ...(values ?? {}) to ...values.
  • Two useless regex escapes in the site generator.
  • String(src) in an invocation error message (a symbol in a template literal would throw).
  • An explicit default in the schema-arbitrary visitor, and a missing active case in a test switch.
  • await act(async () => …) in the React tests.
  • Explicit fields instead of spreading a class instance.
  • An optional parameter instead of = undefined.
• One plugin finding is suppressed: invoke-outcomes-example.ts selects a titled none branch from an onSnapshot resolver. Machine.can does not query snapshot outcomes, so this is a false positive. Check no-conditional-none-branch only on event handlers #257 fixes it by limiting no-conditional-none-branch to on handlers. The suppression can go once Check no-conditional-none-branch only on event handlers #257 merges.

Coordination

• This PR removes four casts in executionPlan.ts, which another in-flight change is also reducing. Whichever merges second should resolve the conflict and rerun pnpm lint.
• When Check no-conditional-none-branch only on event handlers #257 merges, add no-duplicate-sibling-transitions, no-guarded-none, and no-runsync-machine-can to the dogfood override and drop the snapshot suppression. I ran all 14 rules from Check no-conditional-none-branch only on event handlers #257 on these 13 files: 0 findings.
• package.json and pnpm-lock.yaml will need a trivial rebase against the dependency and TS 7 PRs.

Changeset

• Added an empty changeset. The root manifest changed, which the changeset policy requires, but no package release is needed: the source edits only remove type assertions and do not change behavior.
• Not required because this PR does not change a publishable package

Validation

• pnpm check: format, lint, architecture, CI scripts, API docs, site docs, typecheck, build, 901 tests, type tests (246 tests and 1,299 assertions in core, 35 assertions in React), consumer, and all four pack checks passed.
• Fresh clone (git clone + pnpm install --frozen-lockfile + pnpm lint): builds the plugin, then lints 318 files with 119 rules, exit 0. pnpm lint takes about 8–9 s end to end (oxlint alone 5.5–15 s, depending on machine load).
• latest-effect simulation (scripts/prepare-effect-compatibility.mjs, Effect 4.1.1): oxlint and pnpm typecheck both pass.
• Automated type-performance measurement passed or was not required
• Automated runtime- and memory-performance measurement passed or was not required (no runtime change; the casts are erased)

🤖 Generated with Claude Code

[github-actions]

Type performance

Measured with TypeScript 6.0.3 and skipLibCheck=true.

Scenario	Base	PR	Difference
Effect only	55	55	0 (0.0%)
Import effect-machine	55	55	0 (0.0%)
Machine.state (3 children)	2,664	2,664	0 (0.0%)
Machine.make (3 states, 2 events)	8,932	8,932	0 (0.0%)
machine.handle (3 states, 2 transitions)	39,449	39,449	0 (0.0%)
declared transition (10 named branches)	60,402	60,402	0 (0.0%)
atomic construction, guards, and reentry	85,710	85,710	0 (0.0%)
registered invocation (input-taking Effect)	39,819	39,819	0 (0.0%)
machine.handle (depth 24)	194,531	194,531	0 (0.0%)
machine.handle (wide depth 16)	218,238	218,238	0 (0.0%)
machine.handle (parallel/history/choice)	91,642	91,642	0 (0.0%)
machine definition (3 independent implementations)	99,173	99,173	0 (0.0%)
machine exact input/output/error/services	82,941	82,941	0 (0.0%)
execution adapter readiness	71,428	71,428	0 (0.0%)

Marginal instantiations are measured against the matching setup without that API call:

Scenario	Base	PR	Difference
Import effect-machine	0	0	0
Machine.state (3 children)	2,609	2,609	0 (0.0%)
Machine.make (3 states, 2 events)	6,260	6,260	0 (0.0%)
machine.handle (3 states, 2 transitions)	30,517	30,517	0 (0.0%)
declared transition (10 named branches)	51,849	51,849	0 (0.0%)
atomic construction, guards, and reentry	70,370	70,370	0 (0.0%)
registered invocation (input-taking Effect)	32,682	32,682	0 (0.0%)
machine.handle (depth 24)	177,222	177,222	0 (0.0%)
machine.handle (wide depth 16)	201,707	201,707	0 (0.0%)
machine.handle (parallel/history/choice)	79,144	79,144	0 (0.0%)
machine definition (3 independent implementations)	88,492	88,492	0 (0.0%)
machine exact input/output/error/services	70,732	70,732	0 (0.0%)
execution adapter readiness	31,922	31,922	0 (0.0%)

Check times (informational)

Scenario	Base	PR
Effect only	0.03s	0.03s
Import effect-machine	0.03s	0.03s
Machine.state (3 children)	0.09s	0.10s
Machine.make (3 states, 2 events)	0.17s	0.16s
machine.handle (3 states, 2 transitions)	0.31s	0.31s
declared transition (10 named branches)	0.43s	0.41s
atomic construction, guards, and reentry	0.56s	0.52s
registered invocation (input-taking Effect)	0.33s	0.31s
machine.handle (depth 24)	0.72s	0.65s
machine.handle (wide depth 16)	0.72s	0.72s
machine.handle (parallel/history/choice)	0.54s	0.53s
machine definition (3 independent implementations)	0.57s	0.56s
machine exact input/output/error/services	0.52s	0.51s
execution adapter readiness	0.45s	0.44s

Type instantiations are the comparison metric. Check time varies with runner load and is informational only.

[github-actions]

Runtime performance

Median of 5 independent benchmark processes on Intel(R) Xeon(R) 6973P-C with Node v24.22.1.

Pull request baseline

Scenario	Effect Machine
Plan counter transitions	170,455 transitions/s
Drain burst with terminal fence	612,732 increments/s
Drain burst with a change observer	557,123 increments/s
Lookup and send to one child	502,568 increments/s
Start and stop a machine	175,871 machines/s
Start and stop a parent with one child	52,010 families/s
Complete an Effect invocation with tracing enabled	18,619 invocations/s
Complete an Effect invocation with tracing disabled	21,253 invocations/s
Plan transitions through a compound state	149,808 transitions/s
Plan transitions through parallel regions	120,598 transitions/s
Drain burst through a compound state	520,877 events/s
Drain burst through two parallel regions	467,734 events/s
Drain a compound-state burst with a change observer	472,369 events/s

Process runtime reference points

Scenario	Effect Machine
Start and stop a raw generic process	45,612 processes/s
Start and stop a raw compiled process	105,363 processes/s

Memory profile	Effect Machine
Idle machine	1.9 KiB
Raw generic managed process	11.4 KiB
Raw compiled process	3.1 KiB
Two independent idle machines	3.6 KiB
Idle parent with one child	6.0 KiB
Parent with observed child registry	9.6 KiB
Parent with observed invoked child snapshots	6.9 KiB

Effect Machine change from base

Metric	Base	Base variability	PR	PR variability	Difference
Plan counter transitions	169,066 transitions/s	0.3% MAD	170,455 transitions/s	0.5% MAD	+0.8%
Drain burst with terminal fence	631,867 increments/s	1.4% MAD	612,732 increments/s	0.9% MAD	-3.0%
Drain burst with a change observer	579,900 increments/s	0.9% MAD	557,123 increments/s	1.3% MAD	-3.9%
Lookup and send to one child	515,278 increments/s	0.1% MAD	502,568 increments/s	2.4% MAD	-2.5%
Start and stop a machine	177,085 machines/s	0.4% MAD	175,871 machines/s	0.2% MAD	-0.7%
Start and stop a parent with one child	52,832 families/s	0.8% MAD	52,010 families/s	1.3% MAD	-1.6%
Complete an Effect invocation with tracing enabled	18,839 invocations/s	0.6% MAD	18,619 invocations/s	1.2% MAD	-1.2%
Complete an Effect invocation with tracing disabled	21,396 invocations/s	0.6% MAD	21,253 invocations/s	0.9% MAD	-0.7%
Plan transitions through a compound state	149,769 transitions/s	0.4% MAD	149,808 transitions/s	0.7% MAD	+0.0%
Plan transitions through parallel regions	119,024 transitions/s	1.0% MAD	120,598 transitions/s	1.4% MAD	+1.3%
Drain burst through a compound state	507,521 events/s	0.4% MAD	520,877 events/s	0.7% MAD	+2.6%
Drain burst through two parallel regions	470,740 events/s	0.6% MAD	467,734 events/s	0.5% MAD	-0.6%
Drain a compound-state burst with a change observer	468,749 events/s	0.9% MAD	472,369 events/s	2.3% MAD	+0.8%
Idle machine heap per unit	1.9 KiB	0.6% MAD	1.9 KiB	0.2% MAD	-0.3%
Raw generic managed process heap per unit	11.4 KiB	0.0% MAD	11.4 KiB	0.0% MAD	+0.0%
Raw compiled process heap per unit	3.2 KiB	0.1% MAD	3.1 KiB	0.1% MAD	-0.9%
Two independent idle machines heap per unit	3.6 KiB	0.0% MAD	3.6 KiB	0.0% MAD	-0.0%
Idle parent with one child heap per unit	6.0 KiB	0.0% MAD	6.0 KiB	0.1% MAD	+0.1%
Parent with observed child registry heap per unit	9.6 KiB	0.0% MAD	9.6 KiB	0.0% MAD	+0.0%
Parent with observed invoked child snapshots heap per unit	6.9 KiB	0.1% MAD	6.9 KiB	0.2% MAD	-0.0%

Process runtime reference change from base

Metric	Base	Base variability	PR	PR variability	Difference
Start and stop a raw generic process	45,455 processes/s	1.1% MAD	45,612 processes/s	0.5% MAD	+0.3%
Start and stop a raw compiled process	106,236 processes/s	0.6% MAD	105,363 processes/s	0.4% MAD	-0.8%

Regression guard

No large, noise-adjusted throughput or heap regressions detected.

Versions and interpretation

• Effect Machine: 0.46.0

Higher throughput is better; lower heap is better. Variability is the median absolute deviation across independent processes, relative to their median. Small differences on shared GitHub-hosted hardware remain informational; the required guard rejects only large changes beyond the measured noise allowance.