0.7.1g1G5d: close the final NO_COVERAGE cohort (4 UNREACHABLE)

[GionaGranchelli]

Base

cd46f176b20ef7e0e0b2b0e251569214dfc90619 — the squash merge of #463. Worktree branched directly from it, clean tree.

Scope

Exactly the four frozen NO_COVERAGE identities. No widening, no substitution, no production change.

Endpoint

G5D COMPLETE — FINAL G5 NO_COVERAGE COHORT CLOSED. Final classification: 4 UNREACHABLE, 0 KILLED, 0 EQUIVALENT, 0 TOOLING_LIMITATION, 0 UNDETERMINED. Parent accounting: 52 = 36 (G5b) + 12 (G5c) + 4 (G5d). G5 remainder: 0.

identity	owner	mutator	block/index	PC	mutated instruction
564517b2fdab	suspendToolExecution	VoidMethodCall	13/96	pc194	ResultKt.throwOnFailure (resumed path)
66c311cf4f0a	suspendToolExecution	NullReturnVals	12/78	pc163	areturn of COROUTINE_SUSPENDED (suspend exit)
5a7edac3a982	requestApproval	VoidMethodCall	11/92	pc204	ResultKt.throwOnFailure (resumed path)
e52f1c1f6570	requestApproval	NullReturnVals	10/72	pc164	areturn of COROUTINE_SUSPENDED (suspend exit)

Mutator naming: the brief calls the two "NRV" mutants NonReturningVoidMethodCall; the frozen manifest records returns.NullReturnValsMutator ("replaced return value with null"). The frozen identities are authoritative — a label discrepancy, not an identity discrepancy.

Phase A — exact mapping method

PIT reports no bytecode offsets, and reconstructing its block/index counter matched only 4 of 36 calibration identities, so that approach was abandoned rather than tuned until it fit. Each identity is instead bracketed: pitBlock is monotone in pc within a method (verified), and both methods carry in-method calibration points with known pc and known PIT coordinates from the committed G5b/G5c manifests. Every target's block brackets a pc window containing exactly one candidate of the mutated kind.

The source line is useless here and was not relied on: line 142 is the suspendToolExecution function header and line 73 the requestApproval header, so every state-machine instruction inherits it.

Reachability — UNREACHABLE, structurally

All four sit on the suspension protocol of two calls: resolveGovernedSuspension() (line 145) and resolveGovernedIdentity(workflowRunId) (line 82). Neither callee can suspend:

• resolveGovernedSuspension: 50 instructions, 0 getCOROUTINE_SUSPENDED, 0 if_acmpne
• resolveGovernedIdentity-jFE5hGw: 32 instructions, 0 getCOROUTINE_SUSPENDED, 0 if_acmpne; invokes only Continuation.getContext, own-package accessors/getters, Intrinsics.areEqual, concat, an exception ctor
• GovernedRunScope.resolve(context) is a plain non-suspend function

A suspend function returns COROUTINE_SUSPENDED only through the sentinel comparison emitted after a call. With none emitted and no suspending callee, both resolvers always return directly. So the caller's sentinel test is always false and the COROUTINE_SUSPENDED exit cannot execute; and the resumed continuation can only be constructed by a real suspension at that call, so the resumed-path throwOnFailure cannot execute. Nothing else reaches them: no branch or switch entry in either method targets pc194 or pc204.

That is control-flow dominance plus callee analysis — not "I could not find a test".

Test evidence

No tests added — tests are permitted only for proven-reachable identities. Making these execute would require changing production semantics or fabricating a coroutine state the runtime cannot produce. A test existing only to move PIT's coverage counter is what this task forbids.

Phase E — pairs

Both pairs are one compiler-generated suspension: suspendToolExecution block 12 (exit) / block 13 (resumed unwrap); requestApproval block 10 / block 11. Same construct, different instructions, neither dominating. Whether one test would cover both is moot — neither member is reachable. Block/index adjacency was not used as evidence of equivalence.

Diff scope

Tests: none. Production: none. Docs: this report + the four-identity manifest. The measurement narrowing never landed.

Regression gate

Identity-exact campaign running from a fresh worktree at cd46f176 with the proven family-only narrowing; reconciliation (4/4 accounted, 0 identity loss, 0 gain/substitution, 0 new timeouts, 0 regressions of previously KILLED identities, no widening) is appended to the report and manifest on completion.

Limits

UNREACHABLE is a statement about this compiled artifact at this base: if either resolver ever gains a suspension point, these instructions become live and these identities must be re-adjudicated rather than inherited.

Verification: 4/4 mapped with non-null PCs · no churn · spotlessCheck verifyStaticAnalysis verifyChangePolicy -PchangePolicyBase=cd46f176… BUILD SUCCESSFUL. Exact-head CI is the final authority. Not self-merged.

[Copilot]

Copilot review overview

🔵 Needs a closer look

The substance is an unverifiable-from-source bytecode-reachability adjudication (exact PCs, PIT block/index, UNREACHABLE conclusion) that forms mutation-governance audit authority and warrants human/tooling confirmation.

Review effort: Balanced
Findings: None

What changed in this PR

This documentation-only PR closes the final G5 NO_COVERAGE cohort (task TASK-0.7.1g1G5d) by classifying the four remaining frozen mutation-testing identities as UNREACHABLE. It adds an analysis report and a machine-readable manifest that together record the bytecode-exact mapping (Phase A), the control-flow reachability argument (Phase C), the pair analysis (Phase E), and the regression-campaign reconciliation. No production or test code changes; the artifacts extend the mutation-governance audit trail described in AGENTS.md.

Changes:

• Adds a reachability report arguing all four identities are UNREACHABLE because both suspension-protocol resolvers (resolveGovernedSuspension, resolveGovernedIdentity) contain no suspension point, so the COROUTINE_SUSPENDED exits and resumed-path throwOnFailure instructions cannot execute.
• Adds a four-identity JSON manifest with per-identity bytecode PCs, PIT block/index, mutator, evidence, and a regression-campaign reconciliation (918 population, 0 status movements, 0 regressions).
• Records parent accounting (52 = 36 G5b + 12 G5c + 4 G5d; remainder 0) and documents a mutator label discrepancy (brief's "NonReturningVoidMethodCall" vs frozen NullReturnValsMutator).

File	Description
docs/​roadmap/​0.7.0/​TASK-0.7.1g1G5d-NO-COVERAGE-REACHABILITY.md	Narrative report: mapping method, per-identity table, UNREACHABLE reasoning, pair analysis, regression gate, and limits.
docs/​roadmap/​0.7.0/​TASK-0.7.1g1G5d-NO-COVERAGE-4-MANIFEST.json	Structured manifest of the four identities with PCs, calibration windows, classification proof, totals, and regression reconciliation.

I verified that the checkable claims are accurate: the JSON is well-formed; population and duration arithmetic (662+132+62+62=918; 17m13s=1033s) and parent accounting are internally consistent; markdown, JSON, and the PR description agree on identities/PCs/block-index; and the cited source coordinates match production at HEAD (suspendToolExecution at line 142, resolveGovernedSuspension() at 145, requestApproval at 73, resolveGovernedIdentity at 82, and GovernedRunScope.resolve is non-suspend at tramai-core/.../GovernedRunScope.kt:64). I found no objective issues to comment on.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.