Harden statechart modeling and verification

[SandroMaglione]

Summary

• Fix supervisor-owned machine terminalization so initialization and runtime self.stop cannot deadlock, while preserving exactly-once cleanup and terminal publication across stop, completion, and failure races.
• Strengthen public type safety with consistent executable-machine readiness, finite-union event narrowing, exact state-definition validation, and discriminated compiled state-node inspection.
• Expand MachineTest finite models from event-only transitions to canonical event | always | done triggers, including independent stabilization, completion, cycle, resume, generated-model, and managed-runtime differential verification.
• Add adversarial snapshot-codec, activity-lifecycle, process-lifecycle, and compiler-instantiation regression coverage.

MachineTest.FiniteTransition now migrates from { source, event, ... } to { source, trigger: { type: "event", event }, ... }; automatic transitions use { type: "always" } or { type: "done" }.

Changeset

• Added or updated for a library or package-metadata change
• Not required because this PR does not change src/ or package.json

Validation

• pnpm check
• Relevant example checks, when examples changed (not applicable; no examples changed)
• Reviewed the automated type-performance report, when the public TypeScript API or inference changed

[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.defineStates (3 states)	2,902	2,965	+63 (+2.2%)
Machine.make (3 states, 2 events)	8,668	8,731	+63 (+0.7%)
machine.handle (3 states, 2 transitions)	23,474	23,634	+160 (+0.7%)
machine.handle (depth 8)	112,138	113,023	+885 (+0.8%)
machine.handle (depth 12)	124,580	125,661	+1,081 (+0.9%)
machine.handle (depth 16)	139,726	141,003	+1,277 (+0.9%)
machine.handle (depth 24)	178,130	179,799	+1,669 (+0.9%)
machine.handle (wide depth 16)	208,161	210,137	+1,976 (+0.9%)
machine.handle (parallel/history/choice)	—	128,876	new
machine.handle (4 successive calls)	—	129,498	new
machine exact input/output/error/services	—	107,371	new
execution adapter readiness	—	122,890	new

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

Scenario	Base	PR	Difference
Import effect-machine	0	0	0
Machine.defineStates (3 states)	2,847	2,910	+63 (+2.2%)
Machine.make (3 states, 2 events)	5,758	5,758	0 (0.0%)
machine.handle (3 states, 2 transitions)	14,806	14,903	+97 (+0.7%)
machine.handle (depth 8)	105,202	105,546	+344 (+0.3%)
machine.handle (depth 12)	116,640	116,984	+344 (+0.3%)
machine.handle (depth 16)	130,782	131,126	+344 (+0.3%)
machine.handle (depth 24)	167,178	167,522	+344 (+0.2%)
machine.handle (wide depth 16)	197,825	198,169	+344 (+0.2%)
machine.handle (parallel/history/choice)	—	110,514	new
machine.handle (4 successive calls)	—	114,497	new
machine exact input/output/error/services	—	95,819	new
execution adapter readiness	—	96,968	new

Check times (informational)

Scenario	Base	PR
Effect only	0.03s	0.03s
Import effect-machine	0.03s	0.03s
Machine.defineStates (3 states)	0.09s	0.10s
Machine.make (3 states, 2 events)	0.14s	0.14s
machine.handle (3 states, 2 transitions)	0.21s	0.22s
machine.handle (depth 8)	0.50s	0.49s
machine.handle (depth 12)	0.51s	0.53s
machine.handle (depth 16)	0.57s	0.59s
machine.handle (depth 24)	0.64s	0.65s
machine.handle (wide depth 16)	0.70s	0.71s
machine.handle (parallel/history/choice)	—	0.61s
machine.handle (4 successive calls)	—	0.56s
machine exact input/output/error/services	—	0.50s
execution adapter readiness	—	0.55s

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