P12.OP4: Complete score normalization validation — τ<0.95, follow-up bead created

Research validated that both score-based (τ=0.79) and RRF (τ=0.14) merging
fail the 0.95 Kendall tau threshold with skewed shard distributions. Created
follow-up bead miroir-n6v for global-IDF preflight implementation.

Also: add __pycache__/ and tarpaulin-report.json to .gitignore, fix
task_pruner gauge test race.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

.gitignore

@@ -5,3 +5,6 @@
 *.swo
 *~
 .DS_Store
+__pycache__/
+*.pyc
+tarpaulin-report.json

.needle-predispatch-sha

@@ -1 +1 @@
-d2aa49a163694ac47e8a93d6409042eb2b9d99a9
+cec3b819bfeafa596c3ca1d7359ecc602bb7844f

crates/miroir-core/src/task_pruner.rs

@@ -317,6 +317,7 @@ mod tests {
     /// Acceptance: miroir_task_registry_size gauge drops after a prune cycle.
     #[test]
     fn gauge_drops_after_prune() {
+        let _lock = GAUGE_LOCK.lock().unwrap();
         let store = test_store();
         let ten_days_ms: i64 = 10 * 24 * 3600 * 1000;
         let old_time = now() - ten_days_ms;