sigker: fix both clippy findings and arm the lint in CI

.claude/board/TECH_DEBT.md

@@ -1,15 +1,31 @@
-## TD-SIGKER-CLIPPY-RED-ON-BASE-1 (2026-09-04) — OPEN
+## TD-SIGKER-CLIPPY-RED-ON-BASE-1 (2026-09-04) — RESOLVED 2026-09-04
 
 **`crates/sigker` does not pass `cargo clippy --all-targets -- -D warnings`, and
 did not before this session's changes.** Measured by stashing the working tree
 and re-running against the base commit: the failure reproduces identically.
 
-Single site: `crates/sigker/src/signature.rs:133` — `for flat in 0..len` trips
-clippy's needs_range_loop lint: the loop variable is used only to index `level`.
-
-(Anchored on the loop header, not on the lint name. The lint name is not text
-that appears at the cited line, so citing it as the anchor is exactly the decay
-`citation_decay.py` exists to catch — it flagged this entry's first draft.)
+> **⊘ RESOLVED 2026-09-04.** Both findings fixed; sigker now passes
+> `cargo clippy --manifest-path crates/sigker/Cargo.toml --all-targets -- -D warnings`
+> clean, and the CI step below was promoted from tests-only to tests+clippy so
+> it cannot silently return. Two things this entry got WRONG, recorded rather
+> than edited away:
+>
+> 1. **"Single site" was wrong — there were TWO.** `-D warnings` makes clippy
+>    stop at the first error, so `doc list item overindented` in
+>    `examples/depth_scaling.rs` was invisible behind the loop finding and only
+>    appeared once that one was fixed. A findings count taken from a run that
+>    aborted on the first error is a lower bound, never a total.
+> 2. **The citation decayed the moment the fix landed** — deleting the cited
+>    line IS the fix, so the anchor `for flat in 0..len` went absent and
+>    `citation_decay.py` fired on this very entry. Re-anchored below on the
+>    enclosing function, which survives the change. An entry that cites the
+>    code it wants deleted is self-invalidating by construction.
+
+Site: `segment_signature` in `crates/sigker/src/signature.rs:122` — the
+level-fill loop used its loop variable only to index `level`, tripping
+needs_range_loop. Anchored on the function name, not the loop body: a function
+name survives its own body being rewritten, which is exactly the property a
+citation into mutable source needs.
 
 ```rust
 for flat in 0..len {          // clippy: needs_range_loop
@@ -19,8 +35,7 @@ for flat in 0..len {          // clippy: needs_range_loop
 }
 ```
 
-Proposed fix (3 lines, mechanical, deliberately NOT applied here to avoid
-widening a wiring PR):
+Fix, as applied:
 
 ```rust
 for (flat, slot) in level.iter_mut().enumerate() {

.github/workflows/rust-test.yml

@@ -210,15 +210,20 @@ jobs:
       #
       # Verified locally before landing: 62 passed, 0 failed.
       #
-      # Deliberately TESTS ONLY, on the causal-edge precedent above. A
-      # `clippy -D warnings` step would be red on arrival: the crate carries one
-      # pre-existing finding at signature.rs:133 (needs_range_loop), which
-      # reproduces identically on the base commit and is NOT in the
-      # randomized.rs this PR touches. Gating it here would fail this PR for a
-      # defect it did not introduce; it is recorded in TECH_DEBT
-      # TD-SIGKER-CLIPPY-RED-ON-BASE-1 instead, with its patch.
+      # TESTS + CLIPPY. This step was tests-only when first added, because the
+      # crate was clippy-red on arrival (TD-SIGKER-CLIPPY-RED-ON-BASE-1) and
+      # gating it would have failed that PR for a defect it did not introduce.
+      # Both findings are now fixed, so the clippy half is armed here: a lint
+      # fix with no gate decays back, and this crate's whole reason for being
+      # unlinted was that no step reached it.
+      #
+      # Verified locally before landing: 62 tests passed, clippy clean.
       - name: Run sigker tests (workspace-excluded, ndarray sibling)
         run: cargo test --manifest-path crates/sigker/Cargo.toml
+      - name: Lint sigker (workspace-excluded, so root clippy never reaches it)
+        run: |
+          cargo clippy --manifest-path crates/sigker/Cargo.toml \
+            --all-targets -- -D warnings
       # lance-graph-callcenter UNDER `--features query` — the same blind gate as
       # supervisor above, but one level subtler: the crate was not merely
       # untested, it has `default = []`, so even a bare `cargo test` on it would

crates/sigker/examples/depth_scaling.rs

@@ -3,7 +3,8 @@
 //!   - Truncated kernel:    O(d^(2N)) per pair, materializes the signature
 //!   - Goursat-PDE kernel:  O(T₁·T₂) per pair, NEVER materializes the signature
 //!   - Log-signature:       O(d^(2N)) compute (still has Magnus expansion),
-//!                          but storage is dim L_N(d) — 7-13× smaller
+//!     but storage is dim L_N(d) — 7-13× smaller (asymptotic, N≥8; ~2× at
+//!     shallow depth, per `compression_at_shallow_depth_is_far_below_the_headline`)
 //!
 //! Run:
 //!   cargo run --manifest-path crates/sigker/Cargo.toml \

crates/sigker/src/signature.rs

@@ -130,15 +130,15 @@ fn segment_signature(delta: &[f64], depth: usize) -> Signature {
         let mut level = vec![0.0; len];
         // Outer-product expansion of Δ ⊗ ⋯ ⊗ Δ (k times) divided by k!
         // Index mapping: flat_idx = i₁ · d^(k-1) + i₂ · d^(k-2) + … + iₖ.
-        for flat in 0..len {
+        for (flat, slot) in level.iter_mut().enumerate() {
             let mut idx = flat;
             let mut prod = 1.0;
             for _ in 0..k {
                 let ax = idx % dim;
                 idx /= dim;
                 prod *= delta[ax];
             }
-            level[flat] = prod / factorial;
+            *slot = prod / factorial;
         }
         levels.push(level);
     }