chore: workspace-wide cleanup — clippy::pedantic baseline + auto-fix

cut PR v0.18.0 전 마지막 정리. 사용자 요청: "전체 코드베이스를 깔끔하고 알아보기 쉽게".

## Workspace lints

- `Cargo.toml` 의 `[workspace.lints.clippy]` 에 `pedantic = "warn"` (priority -1) + 의도적 allow-list 추가:
  - cast_possible_truncation / cast_possible_wrap / cast_sign_loss / cast_precision_loss — ONNX i64 / hash modular reduction 등 의도적 truncation.
  - doc_markdown / missing_errors_doc / missing_panics_doc — cosmetic doc style.
  - too_many_lines / module_name_repetitions / must_use_candidate / needless_pass_by_value / manual_let_else / items_after_statements / similar_names — informational only.
  - format_collect / match_wildcard_for_single_variants / trivially_copy_pass_by_ref / unnecessary_wraps — intentional patterns (exhaustive match, future Result variants 등).
  - default_trait_access — `Foo::default()` 가 idiomatic.
  - float_cmp — NLI / RRF score 의 explicit threshold 비교 의도.
  - struct_excessive_bools / case_sensitive_file_extension_comparisons / naive_bytecount / ignore_without_reason — domain-specific 의도.
  - format_push_string / return_self_not_must_use / match_same_arms — builder / wire-label / hot-path 패턴 보존.
  - needless_continue / used_underscore_binding / nonminimal_bool / unreadable_literal / many_single_char_names / doc_link_with_quotes / assigning_clones / collapsible_str_replace / trivial_regex / elidable_lifetime_names / range_plus_one / explicit_iter_loop / implicit_hasher / ref_option — remaining low-value style.
- 각 24 crate `Cargo.toml` 에 `[lints] workspace = true` 추가.

## Auto-fix

`cargo clippy --workspace --all-targets --fix` 적용 — 128 files changed, 552 insertions / 472 deletions. 주로:
- uninlined_format_args (~18): `format!("{}", x)` → `format!("{x}")`.
- redundant_closure_for_method_calls (~33): `.map(|x| x.foo())` → `.map(T::foo)`.
- 그 외 mechanical refactor.

## 검증

- `cargo clippy --workspace --all-targets -j 1 -- -D warnings` clean (pedantic + 모든 lint group).
- `cargo test --workspace --no-fail-fast -j 1` — **1293 tests pass + 1 pre-existing flaky fail** (`kebab-mcp::tools_call_ask_multi_hop::ask_tool_routes_multi_hop_true_to_decompose_first`, HOTFIX candidate, cleanup 무관). 회귀 0.

Wire 영향: 없음.
Behavior 영향: 없음 (mechanical refactor only).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

crates/kebab-eval/Cargo.toml

@@ -28,3 +28,6 @@ uuid            = { workspace = true }
 [dev-dependencies]
 tempfile        = { workspace = true }
 rusqlite        = { workspace = true }
+
+[lints]
+workspace = true

crates/kebab-eval/src/compare.rs

@@ -260,8 +260,8 @@ pub fn render_report_md(report: &CompareReport) -> String {
             "| {} | {} | {} | {} | {} |",
             c.query_id,
             comparison_kind_label(c.kind),
-            c.a_hit_rank.map(|r| r.to_string()).unwrap_or_else(|| "—".into()),
-            c.b_hit_rank.map(|r| r.to_string()).unwrap_or_else(|| "—".into()),
+            c.a_hit_rank.map_or_else(|| "—".into(), |r| r.to_string()),
+            c.b_hit_rank.map_or_else(|| "—".into(), |r| r.to_string()),
             c.note.as_deref().unwrap_or(""),
         );
     }
@@ -308,7 +308,7 @@ fn extract_chunker_version(snapshot_json: &str) -> Option<String> {
     let v: serde_json::Value = serde_json::from_str(snapshot_json).ok()?;
     v.get("chunker_version")
         .and_then(|x| x.as_str())
-        .map(|s| s.to_owned())
+        .map(std::borrow::ToOwned::to_owned)
 }
 
 fn parse_results(
@@ -402,8 +402,7 @@ fn classify(
             // so refusal-flow queries (no expected_*) don't appear as
             // regressions.
             let has_expected = gq
-                .map(|g| !g.expected_chunk_ids.is_empty() || !g.expected_doc_ids.is_empty())
-                .unwrap_or(false);
+                .is_some_and(|g| !g.expected_chunk_ids.is_empty() || !g.expected_doc_ids.is_empty());
             if has_expected {
                 (ComparisonKind::Regression, Some("hit→miss".into()))
             } else {
@@ -426,7 +425,7 @@ fn build_deltas(
         if a.is_nan() || b.is_nan() {
             serde_json::Value::Null
         } else {
-            serde_json::Value::from((b - a) as f64)
+            serde_json::Value::from(f64::from(b - a))
         }
     }
     let mut hit = serde_json::Map::new();

crates/kebab-eval/src/metrics.rs

@@ -270,7 +270,21 @@ pub(crate) fn aggregate_from_rows(
         // recall@k_doc (doc-level, requires non-empty expected_doc_ids
         // and `>0` is the "should retrieve" condition; refusal queries
         // (`expected_doc_ids = []`) are excluded by spec).
-        if !gq.expected_doc_ids.is_empty() {
+        if gq.expected_doc_ids.is_empty() {
+            // refusal_correctness: golden marks "should refuse" via empty
+            // expected_doc_ids. We can only judge this on RAG runs — a
+            // lexical-only run produces no Answer, so "refusal" is
+            // undefined. Excluding such queries from the denominator
+            // (rather than counting them as failures) keeps the metric
+            // honest: a search-only run reports refusal_correctness as
+            // NaN/null, not 0.0.
+            if let Some(ans) = &qr.answer {
+                refusal_denom += 1;
+                if !ans.grounded {
+                    refusal_num += 1;
+                }
+            }
+        } else {
             let expected_docs: HashSet<&DocumentId> = gq.expected_doc_ids.iter().collect();
             for k in TOP_K_VARIANTS {
                 let entry = recall_at_k_doc.get_mut(k).expect("init");
@@ -285,20 +299,6 @@ pub(crate) fn aggregate_from_rows(
                 let frac = covered as f64 / expected_docs.len() as f64;
                 entry.0 += frac;
             }
-        } else {
-            // refusal_correctness: golden marks "should refuse" via empty
-            // expected_doc_ids. We can only judge this on RAG runs — a
-            // lexical-only run produces no Answer, so "refusal" is
-            // undefined. Excluding such queries from the denominator
-            // (rather than counting them as failures) keeps the metric
-            // honest: a search-only run reports refusal_correctness as
-            // NaN/null, not 0.0.
-            if let Some(ans) = &qr.answer {
-                refusal_denom += 1;
-                if !ans.grounded {
-                    refusal_num += 1;
-                }
-            }
         }
 
         // groundedness + citation_coverage (only meaningful with RAG

crates/kebab-eval/tests/metrics_and_compare.rs

@@ -143,7 +143,7 @@ fn env_guard() -> std::sync::MutexGuard<'static, ()> {
     static M: OnceLock<Mutex<()>> = OnceLock::new();
     M.get_or_init(|| Mutex::new(()))
         .lock()
-        .unwrap_or_else(|e| e.into_inner())
+        .unwrap_or_else(std::sync::PoisonError::into_inner)
 }
 
 #[test]

crates/kebab-eval/tests/runner.rs

@@ -147,7 +147,7 @@ fn lexical_opts() -> EvalRunOpts {
 /// guard must outlive the call so concurrent tests don't reset the
 /// var mid-run.
 fn run_with_golden<F: FnOnce() -> R, R>(yaml: &Path, f: F) -> R {
-    let _g = GOLDEN_ENV_LOCK.lock().unwrap_or_else(|p| p.into_inner());
+    let _g = GOLDEN_ENV_LOCK.lock().unwrap_or_else(std::sync::PoisonError::into_inner);
     // SAFETY: `KEBAB_EVAL_GOLDEN` is a benign env var; the GOLDEN_ENV_LOCK
     // serializes mutations so concurrent tests don't race.
     unsafe {