Finding
kanon gate's lint step (strict-warnings, whole-tree) fails on 12 WORKFLOW/unwired-dead-code-untracked ERRORs: #[allow(dead_code)] suppressions across the workspace whose reason strings carry no tracking-issue ref (#NNN) or benign marker (RAII/held/keeps-alive/guard/test-fixture/schema-contract). Because the gate lints the whole tree, these block every harmonia PR, not just ones touching the flagged files.
Evidence
Gate lint on 2026-07-18 (fresh kanon binary): 12 errors, e.g.
crates/akouo-core/src/engine.rs:101, :923crates/eksetasis/src/cf_bypass/byparr.rs:51crates/epignosis/src/providers/audnexus.rs:73, :79, :85crates/epignosis/src/resolver.rs:61- (+ more across the workspace)
6 findings are already baseline-suppressed; these 12 are not.
Why this matters
A mechanical fleet-hygiene change (the zetesis→eksetasis crate rename, kanon#2402) is committed, compiles clean, and gate-blocked solely by these pre-existing suppressions — they are unrelated to the change but fail its gate. Any harmonia PR hits the same wall.
Desired correction
For each flagged suppression: add a topically-relevant tracking-issue ref (#NNN — file per-feature wiring issues where the code is a staged-but-unwired feature) or a benign marker where the dead code is genuinely benign (RAII guard, schema contract, etc.). Where the suppression represents genuinely dead code, remove it and the code. Do not blanket-baseline without triage — the burndown is the point.
Done when: kanon lint reports zero WORKFLOW/unwired-dead-code-untracked errors on harmonia's tree, so PRs gate on their own merits.
Finding
kanon gate's lint step (strict-warnings, whole-tree) fails on 12WORKFLOW/unwired-dead-code-untrackedERRORs:#[allow(dead_code)]suppressions across the workspace whose reason strings carry no tracking-issue ref (#NNN) or benign marker (RAII/held/keeps-alive/guard/test-fixture/schema-contract). Because the gate lints the whole tree, these block every harmonia PR, not just ones touching the flagged files.Evidence
Gate lint on 2026-07-18 (fresh kanon binary): 12 errors, e.g.
crates/akouo-core/src/engine.rs:101, :923crates/eksetasis/src/cf_bypass/byparr.rs:51crates/epignosis/src/providers/audnexus.rs:73, :79, :85crates/epignosis/src/resolver.rs:616 findings are already baseline-suppressed; these 12 are not.
Why this matters
A mechanical fleet-hygiene change (the
zetesis→eksetasiscrate rename, kanon#2402) is committed, compiles clean, and gate-blocked solely by these pre-existing suppressions — they are unrelated to the change but fail its gate. Any harmonia PR hits the same wall.Desired correction
For each flagged suppression: add a topically-relevant tracking-issue ref (
#NNN— file per-feature wiring issues where the code is a staged-but-unwired feature) or a benign marker where the dead code is genuinely benign (RAII guard, schema contract, etc.). Where the suppression represents genuinely dead code, remove it and the code. Do not blanket-baseline without triage — the burndown is the point.Done when:
kanon lintreports zeroWORKFLOW/unwired-dead-code-untrackederrors on harmonia's tree, so PRs gate on their own merits.