[martin-kolinek]

Summary

• Address all six mutation survivors across four source sites from the actual post-merge scheduled run.
• Precisely exclude the equivalent default-value replacement and the native canonicalization forwarding mutation, retaining scripted result/error-policy coverage.
• Split platform-specific path identity and transient-error classification into cfg-gated function items so Gamma 0.4.0 does not mutate disabled tail expressions.
• Leave generated workflows, Gamma 0.4.0, the 100% score threshold, and the zero-flaky gate unchanged.

Verification

The source scheduled run exposed only the two exhaustive mutation-gate failures: https://github.com/microsoft/ox-tools/actions/runs/42976692409

The actual full anvil-scheduled workflow passed at this PR head, ac54232: https://github.com/microsoft/ox-tools/actions/runs/42996589390

• All 14 validation jobs passed; the failure publisher correctly skipped.
• Linux: 59,858/59,858 mutants killed. Windows: 59,598/59,598 killed. Zero survivors, timeouts, OOMs, or uncovered mutants on either platform.
• Regular tests, advisories, and runtime analysis passed on all four platforms. Feature powersets and benchmark compilation passed on both exhaustive platforms.
• Total elapsed approximately 2h26m. This was the real scheduled workflow dispatched on the branch, not the former protected diagnostic harness.

Focused regressions passed on Linux and Windows; all-feature library Clippy passed on Windows. Broad local library runs were not green due to developer-host Cargo configuration, Windows symlink privilege, and WSL-mounted filesystem limitations. The complete hosted workflow is the acceptance evidence.

[Copilot]

🟢 Approval recommended

The supplied review reports no active issues with the targeted mutation-testing changes.

No open issues 🎉

What changed in this PR

This PR addresses scheduled mutation-testing gaps in cargo-gamma-lib without changing runtime behavior or relaxing the mutation-score and zero-flaky gates.

Changes:

• Splits platform-specific path comparison and transient-error classification into separately cfg-gated functions.
• Adds targeted mutation-skip comments for an equivalent default replacement and the native canonicalization adapter, retaining scripted policy coverage.
• Leaves generated workflows and Gamma 0.4.0 unchanged.

File	Description
crates/​cargo-gamma-lib/​src/​exec/​workspace.rs	Separates Windows and non-Windows path-identity functions.
crates/​cargo-gamma-lib/​src/​exec/​verdict.rs	Separates Unix and non-Unix transient-error classifiers.
crates/​cargo-gamma-lib/​src/​exec/​testing_state.rs	Adds a targeted skip comment for native canonicalization forwarding.
crates/​cargo-gamma-lib/​src/​cfg/​cfgs.rs	Adds a targeted skip comment for an equivalent default-value mutation.

_{Review effort: Balanced}

[Codecov Comments Bot (codecov-commenter)]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.5%. Comparing base (a093f71) to head (ac54232).

Additional details and impacted files

@@          Coverage Diff          @@
##            main    #359   +/-   ##
=====================================
  Coverage   99.5%   99.5%           
=====================================
  Files        351     351           
  Lines      64349   64353    +4     
=====================================
+ Hits       64050   64054    +4     
  Misses       299     299           

Flag	Coverage Δ
linux	99.4% <100.0%> (ø)
linux-arm	99.4% <100.0%> (ø)
scheduled	99.5% <100.0%> (+<0.1%)	⬆️
windows	99.5% <100.0%> (-0.1%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.