[avrabe]

#522's per-PR gate ran cargo mutants --in-diff under the workflow-wide
RUSTFLAGS "-D warnings". A mutant that replaces a function body leaves
its parameters unused, and -D warnings makes that a compile error
("unused variable: inst"). On #517 all 9 selected mutants were
therefore UNVIABLE, nothing was caught, and the gate's "none caught"
rule failed the PR with no test ever having run. Same diff measured
locally under RUSTFLAGS="-D warnings": 9 unviable, reproduced exactly.

--cap-lints=true makes cargo-mutants cap lint levels for mutated
builds. Same diff, same RUSTFLAGS, with the flag: 9 selected, 6 caught,
3 unviable, 0 missed, matching the local run without -D warnings.

Only the in-diff step changes. The full-crate run keeps -D warnings,
because its 292 ratchet was measured under it and capping would change
which mutants compile.

The "none caught → FAIL" rule did its job here. It refused to read "0
missed" from a run in which no test executed.

Refs #400, #496

🤖 Generated with Claude Code

https://claude.ai/code/session_01BnBWxgVEUPdsBTLNKtrA6Z

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[github-actions]

Rivet verification gate

✅ 20/20 passed

	count
Passed	20
Failed	0
Skipped (no steps)	0

Filter: (and (= type "feature") (or (has-tag "v093") (has-tag "v0100")))

Failed artifacts

(none)

_{Updated automatically by tools/post_verification_comment.py. Source of truth: artifacts/verification.yaml.}