Upgrade to Effect RC.116 and native property testing

[SandroMaglione]

Summary

Upgrade Effect and its companion packages from RC.112 to RC.116. The testing API now accepts and returns native Effect Arbitrary values, including the newly released array shrinker:

// Before
FastCheck.array(commandArbitrary, { maxLength: 50 })
// After
Arbitrary.array(commandArbitrary, { maxLength: 50 })

Migrate generator composition, schema annotations, sampling, property options, replay, CLI flags, and schema getters. Command-sequence regression coverage verifies removal of unrelated commands and exact replay. The new generator exposed a reference-model error in ancestor initial-choice lifecycle paths; fix its boundary calculation and cover both reenter settings without changing the planner/runtime.

Document migration of custom generators and replay tokens. Extend the existing fresh-release exceptions to the exact Effect companion packages required by this upgrade.

Changeset

• Added or updated for a library or package-metadata change
• Not required because this PR does not change a publishable package

Minor changeset for the public testing generator type change. The fixed release group keeps all four packages synchronized.

Validation

• pnpm check — 778 tests, core and React type assertions, strict consumer compilation, and all package checks
• Automated type-performance measurement passed or was not required
• Automated runtime- and memory-performance measurement passed or was not required

pnpm perf:types and pnpm perf:runtime passed locally. Final-head CI also passed all correctness, packaging, and performance checks.

CI type instantiations decreased across measured machine scenarios. Five alternating runtime runs per revision found no noise-adjusted throughput or heap regressions; measured machine throughput improved by roughly 5–16%. Reviewed both comparison reports for head 403acb3 against base dde2b81.

• Type comparison workflow
• Runtime and memory comparison workflow

[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)	3,521	2,664	-857 (-24.3%)
Machine.make (3 states, 2 events)	9,488	8,632	-856 (-9.0%)
machine.handle (3 states, 2 transitions)	49,981	49,125	-856 (-1.7%)
declared transition (10 named branches)	63,038	62,183	-855 (-1.4%)
atomic construction, guards, and reentry	88,529	87,731	-798 (-0.9%)
registered invocation (input-taking Effect)	49,785	48,929	-856 (-1.7%)
machine.handle (depth 24)	197,380	196,525	-855 (-0.4%)
machine.handle (wide depth 16)	221,123	220,268	-855 (-0.4%)
machine.handle (parallel/history/choice)	95,104	94,249	-855 (-0.9%)
machine definition (3 independent implementations)	100,566	99,583	-983 (-1.0%)
machine exact input/output/error/services	85,901	84,918	-983 (-1.1%)
execution adapter readiness	73,376	72,521	-855 (-1.2%)

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)	3,466	2,609	-857 (-24.7%)
Machine.make (3 states, 2 events)	5,959	5,960	+1 (+0.0%)
machine.handle (3 states, 2 transitions)	40,493	40,493	0 (0.0%)
declared transition (10 named branches)	53,930	53,930	0 (0.0%)
atomic construction, guards, and reentry	72,720	72,720	0 (0.0%)
registered invocation (input-taking Effect)	42,088	42,088	0 (0.0%)
machine.handle (depth 24)	179,529	179,530	+1 (+0.0%)
machine.handle (wide depth 16)	204,050	204,051	+1 (+0.0%)
machine.handle (parallel/history/choice)	82,064	82,065	+1 (+0.0%)
machine definition (3 independent implementations)	89,344	89,216	-128 (-0.1%)
machine exact input/output/error/services	73,166	73,038	-128 (-0.2%)
execution adapter readiness	23,179	23,180	+1 (+0.0%)

Check times (informational)

Scenario	Base	PR
Effect only	0.05s	0.03s
Import effect-machine	0.03s	0.03s
Machine.state (3 children)	0.11s	0.09s
Machine.make (3 states, 2 events)	0.16s	0.14s
machine.handle (3 states, 2 transitions)	0.36s	0.34s
declared transition (10 named branches)	0.43s	0.40s
atomic construction, guards, and reentry	0.56s	0.50s
registered invocation (input-taking Effect)	0.35s	0.34s
machine.handle (depth 24)	0.66s	0.65s
machine.handle (wide depth 16)	0.72s	0.71s
machine.handle (parallel/history/choice)	0.52s	0.51s
machine definition (3 independent implementations)	0.57s	0.55s
machine exact input/output/error/services	0.51s	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 AMD EPYC 7763 64-Core Processor with Node v24.20.0.

Pull request baseline

Scenario	Effect Machine
Plan counter transitions	113,343 transitions/s
Drain burst with terminal fence	394,661 increments/s
Drain burst with a change observer	366,501 increments/s
Lookup and send to one child	323,062 increments/s
Start and stop a machine	130,141 machines/s
Start and stop a parent with one child	26,391 families/s
Complete an Effect invocation with tracing enabled	7,369 invocations/s
Complete an Effect invocation with tracing disabled	9,128 invocations/s
Plan transitions through a compound state	101,464 transitions/s
Plan transitions through parallel regions	80,161 transitions/s
Drain burst through a compound state	340,046 events/s
Drain burst through two parallel regions	312,937 events/s
Drain a compound-state burst with a change observer	315,051 events/s

Process runtime reference points

Scenario	Effect Machine
Start and stop a raw generic process	21,383 processes/s
Start and stop a raw compiled process	71,963 processes/s

Memory profile	Effect Machine
Idle machine	1.8 KiB
Raw generic managed process	11.6 KiB
Raw compiled process	3.0 KiB
Two independent idle machines	3.4 KiB
Idle parent with one child	5.8 KiB
Parent with observed child registry	9.3 KiB
Parent with observed invoked child snapshots	6.4 KiB

Effect Machine change from base

Metric	Base	Base variability	PR	PR variability	Difference
Plan counter transitions	103,921 transitions/s	2.5% MAD	113,343 transitions/s	1.7% MAD	+9.1%
Drain burst with terminal fence	368,526 increments/s	0.4% MAD	394,661 increments/s	0.6% MAD	+7.1%
Drain burst with a change observer	348,252 increments/s	0.4% MAD	366,501 increments/s	0.8% MAD	+5.2%
Lookup and send to one child	303,558 increments/s	0.8% MAD	323,062 increments/s	0.9% MAD	+6.4%
Start and stop a machine	111,894 machines/s	0.8% MAD	130,141 machines/s	0.8% MAD	+16.3%
Start and stop a parent with one child	23,299 families/s	0.9% MAD	26,391 families/s	1.6% MAD	+13.3%
Complete an Effect invocation with tracing enabled	6,916 invocations/s	0.3% MAD	7,369 invocations/s	1.4% MAD	+6.5%
Complete an Effect invocation with tracing disabled	8,410 invocations/s	0.6% MAD	9,128 invocations/s	0.7% MAD	+8.5%
Plan transitions through a compound state	92,414 transitions/s	1.8% MAD	101,464 transitions/s	1.4% MAD	+9.8%
Plan transitions through parallel regions	75,595 transitions/s	1.3% MAD	80,161 transitions/s	1.9% MAD	+6.0%
Drain burst through a compound state	300,805 events/s	2.0% MAD	340,046 events/s	1.1% MAD	+13.0%
Drain burst through two parallel regions	280,639 events/s	1.8% MAD	312,937 events/s	0.9% MAD	+11.5%
Drain a compound-state burst with a change observer	284,901 events/s	1.6% MAD	315,051 events/s	0.5% MAD	+10.6%
Idle machine heap per unit	1.9 KiB	0.7% MAD	1.8 KiB	1.0% MAD	-3.9%
Raw generic managed process heap per unit	13.9 KiB	0.0% MAD	11.6 KiB	0.0% MAD	-16.6%
Raw compiled process heap per unit	3.2 KiB	0.4% MAD	3.0 KiB	0.4% MAD	-4.5%
Two independent idle machines heap per unit	3.5 KiB	0.0% MAD	3.4 KiB	0.0% MAD	-2.9%
Idle parent with one child heap per unit	6.0 KiB	0.0% MAD	5.8 KiB	0.0% MAD	-1.9%
Parent with observed child registry heap per unit	10.0 KiB	0.0% MAD	9.3 KiB	0.0% MAD	-7.2%
Parent with observed invoked child snapshots heap per unit	6.5 KiB	0.1% MAD	6.4 KiB	0.2% MAD	-1.9%

Process runtime reference change from base

Metric	Base	Base variability	PR	PR variability	Difference
Start and stop a raw generic process	17,685 processes/s	1.5% MAD	21,383 processes/s	1.5% MAD	+20.9%
Start and stop a raw compiled process	60,713 processes/s	0.5% MAD	71,963 processes/s	0.5% MAD	+18.5%

Regression guard

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

Versions and interpretation

• Effect Machine: 0.37.0

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.