Fix initial choice state decoding

[SandroMaglione]

Summary

• Decode schema-backed state constructions before an initial choice resolver reads containingState or ancestors.
• Preserve encoded target values for normal entry planning while giving choice selection the same decoded state view as ordinary active-state entry.
• Add public nested-state coverage and forced generic-versus-compiled invoke/retry lifecycle coverage.

Changeset

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

Patch changeset for @typeonce/effect-machine.

Validation

• pnpm check
• Automated type-performance measurement passed or was not required
• Automated runtime- and memory-performance measurement passed or was not required

Focused validation: 43 choice and strategy-differential tests passed.

Runtime performance: local pnpm perf:runtime smoke passed; the PR workflow base-versus-PR comparison remains authoritative. Type-performance validation is not required because this change does not alter the public TypeScript API or inference.

[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.states (3 states)	3,039	3,039	0 (0.0%)
Machine.make (3 states, 2 events)	10,787	10,787	0 (0.0%)
machine.handle (3 states, 2 transitions)	28,551	28,551	0 (0.0%)
fluent transition (10 named branches)	122,775	122,775	0 (0.0%)
fluent invocation (state-dependent Effect)	108,429	108,429	0 (0.0%)
machine.handle (depth 24)	207,970	207,970	0 (0.0%)
machine.handle (wide depth 16)	236,332	236,332	0 (0.0%)
machine.handle (parallel/history/choice)	144,812	144,812	0 (0.0%)
machine definition (3 independent implementations)	143,617	143,617	0 (0.0%)
machine exact input/output/error/services	129,873	129,873	0 (0.0%)
execution adapter readiness	136,469	136,469	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.states (3 states)	2,984	2,984	0 (0.0%)
Machine.make (3 states, 2 events)	7,740	7,740	0 (0.0%)
machine.handle (3 states, 2 transitions)	17,764	17,764	0 (0.0%)
fluent transition (10 named branches)	112,498	112,498	0 (0.0%)
fluent invocation (state-dependent Effect)	99,018	99,018	0 (0.0%)
machine.handle (depth 24)	187,474	187,474	0 (0.0%)
machine.handle (wide depth 16)	217,043	217,043	0 (0.0%)
machine.handle (parallel/history/choice)	124,167	124,167	0 (0.0%)
machine definition (3 independent implementations)	127,576	127,576	0 (0.0%)
machine exact input/output/error/services	115,217	115,217	0 (0.0%)
execution adapter readiness	106,299	106,299	0 (0.0%)

Check times (informational)

Scenario	Base	PR
Effect only	0.03s	0.03s
Import effect-machine	0.03s	0.03s
Machine.states (3 states)	0.10s	0.10s
Machine.make (3 states, 2 events)	0.16s	0.15s
machine.handle (3 states, 2 transitions)	0.25s	0.24s
fluent transition (10 named branches)	0.54s	0.56s
fluent invocation (state-dependent Effect)	0.56s	0.54s
machine.handle (depth 24)	0.78s	0.77s
machine.handle (wide depth 16)	0.79s	0.80s
machine.handle (parallel/history/choice)	0.65s	0.63s
machine definition (3 independent implementations)	0.65s	0.64s
machine exact input/output/error/services	0.60s	0.64s
execution adapter readiness	0.59s	0.59s

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 v24.19.0.

Pull request baseline

Scenario	Effect Machine
Plan counter transitions	168,096 transitions/s
Drain burst with terminal fence	553,840 increments/s
Drain burst with a change observer	509,978 increments/s
Lookup and send to one child	468,014 increments/s
Start and stop a machine	209,336 machines/s
Start and stop a parent with one child	58,197 families/s
Plan transitions through a compound state	148,912 transitions/s
Plan transitions through parallel regions	119,960 transitions/s
Drain burst through a compound state	436,118 events/s
Drain burst through two parallel regions	412,838 events/s
Drain a compound-state burst with a change observer	405,940 events/s

Process runtime reference points

Scenario	Effect Machine
Start and stop a raw generic process	38,971 processes/s
Start and stop a raw compiled process	90,835 processes/s

Memory profile	Effect Machine
Idle machine	1.7 KiB
Raw generic managed process	13.9 KiB
Raw compiled process	3.2 KiB
Two independent idle machines	3.3 KiB
Idle parent with one child	5.7 KiB
Parent with observed child registry	9.8 KiB
Parent with observed invoked child snapshots	6.3 KiB

Effect Machine change from base

Metric	Base	Base variability	PR	PR variability	Difference
Plan counter transitions	170,741 transitions/s	2.8% MAD	168,096 transitions/s	1.5% MAD	-1.5%
Drain burst with terminal fence	596,474 increments/s	9.0% MAD	553,840 increments/s	4.9% MAD	-7.1%
Drain burst with a change observer	545,473 increments/s	7.4% MAD	509,978 increments/s	4.4% MAD	-6.5%
Lookup and send to one child	497,357 increments/s	5.9% MAD	468,014 increments/s	3.6% MAD	-5.9%
Start and stop a machine	207,943 machines/s	1.8% MAD	209,336 machines/s	1.6% MAD	+0.7%
Start and stop a parent with one child	57,740 families/s	5.4% MAD	58,197 families/s	2.2% MAD	+0.8%
Plan transitions through a compound state	155,595 transitions/s	0.8% MAD	148,912 transitions/s	2.0% MAD	-4.3%
Plan transitions through parallel regions	124,769 transitions/s	0.9% MAD	119,960 transitions/s	2.3% MAD	-3.9%
Drain burst through a compound state	450,933 events/s	2.9% MAD	436,118 events/s	2.8% MAD	-3.3%
Drain burst through two parallel regions	435,682 events/s	2.5% MAD	412,838 events/s	2.5% MAD	-5.2%
Drain a compound-state burst with a change observer	421,284 events/s	2.8% MAD	405,940 events/s	1.1% MAD	-3.6%
Idle machine heap per unit	1.7 KiB	0.2% MAD	1.7 KiB	0.3% MAD	+0.1%
Raw generic managed process heap per unit	13.9 KiB	0.0% MAD	13.9 KiB	0.0% MAD	+0.0%
Raw compiled process heap per unit	3.2 KiB	0.3% MAD	3.2 KiB	0.2% MAD	-0.1%
Two independent idle machines heap per unit	3.3 KiB	0.0% MAD	3.3 KiB	0.0% MAD	-0.0%
Idle parent with one child heap per unit	5.7 KiB	0.0% MAD	5.7 KiB	0.0% MAD	0.0%
Parent with observed child registry heap per unit	9.8 KiB	0.0% MAD	9.8 KiB	0.0% MAD	-0.0%
Parent with observed invoked child snapshots heap per unit	6.3 KiB	0.1% MAD	6.3 KiB	0.1% MAD	+0.1%

Process runtime reference change from base

Metric	Base	Base variability	PR	PR variability	Difference
Start and stop a raw generic process	38,609 processes/s	1.5% MAD	38,971 processes/s	2.0% MAD	+0.9%
Start and stop a raw compiled process	92,064 processes/s	2.1% MAD	90,835 processes/s	2.6% MAD	-1.3%

Regression guard

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

Versions and interpretation

• Effect Machine: 0.31.1

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.