fix: resolve integration test compilation + CI errors

Test compilation fixes (8 integration test files):
  1. Ambiguous float types — added f32/f64 annotations
  2. chrono API — replaced with_hour() with date_naive().and_hms_opt()
  3. Missing dev-dependencies — added sqlx + base64
  4. Generic parse — wrapped f32 comparison in parens
  5. Incorrect assertion — 3^5=243 > 100, changed nodes to 1000

CI fixes:
  6. Missing benchmark fixtures — created 3 files in fixtures/benchmarks/
  7. clippy absurd_extreme_comparisons — usize >= 0 always true
  8. authentik_jwt test — Option<SystemTime> type mismatch
  9. http_server tests — removed broken RBAC test module (types deleted)

Result: cargo build --all clean, cargo test --all --lib passes

fixtures/benchmarks/json-output.json

@@ -0,0 +1,28 @@
+{
+  "results": [
+    {
+      "id": "chunk-abc123",
+      "level": "L1",
+      "score": 0.95,
+      "text": "Kubernetes uses port 8080 for API server",
+      "source": "transcript://session-001"
+    },
+    {
+      "id": "chunk-def456",
+      "level": "L2",
+      "score": 0.87,
+      "text": "Common debugging pattern for CrashLoopBackOff pods",
+      "source": "transcript://session-002"
+    },
+    {
+      "id": "chunk-ghi789",
+      "level": "R",
+      "score": 0.72,
+      "text": "See kubectl troubleshooting guide section 3.2",
+      "source": "obsidian://poimen-vault/kubectl.md"
+    }
+  ],
+  "total_hits": 127,
+  "search_time_ms": 145,
+  "query": "fix kubernetes port conflict"
+}

fixtures/benchmarks/markdown-docs.txt

@@ -0,0 +1,33 @@
+# Kubernetes Troubleshooting Guide
+
+## Port Conflicts
+
+When a port conflict occurs on port 8080, check for existing services:
+
+```bash
+kubectl get svc --all-namespaces | grep 8080
+```
+
+### Common Causes
+
+1. Multiple services binding to same NodePort
+2. Host network pods conflicting with node services
+3. Ingress controller port overlap
+
+## CrashLoopBackOff
+
+Pods enter CrashLoopBackOff when the container exits repeatedly.
+
+### Diagnosis Steps
+
+1. Check pod logs: `kubectl logs <pod> --previous`
+2. Check events: `kubectl describe pod <pod>`
+3. Check resource limits: memory/CPU constraints
+4. Check liveness probes: incorrect health check paths
+
+### Resolution
+
+- Increase memory limits if OOMKilled
+- Fix application startup errors
+- Adjust probe timing (initialDelaySeconds)
+- Check environment variable configuration

fixtures/benchmarks/mixed-logs.txt

@@ -0,0 +1,16 @@
+2025-01-15T10:00:00Z INFO  Starting service on port 8080
+2025-01-15T10:00:01Z DEBUG Database connection pool initialized (max=20)
+2025-01-15T10:00:02Z INFO  Health check endpoint ready at /health
+2025-01-15T10:00:05Z WARN  High memory usage detected: 85% of 512Mi limit
+2025-01-15T10:00:10Z ERROR Connection refused: temporal-frontend:7233
+2025-01-15T10:00:15Z INFO  Retry attempt 1/3 for temporal connection
+2025-01-15T10:00:20Z INFO  Connected to temporal-frontend.temporal.svc.cluster.local:7233
+2025-01-15T10:00:25Z DEBUG Worker registered on task queue: poimen-taskqueue
+2025-01-15T10:00:30Z INFO  Processing ingest request: project=poimen source=transcript://session-001
+2025-01-15T10:00:31Z DEBUG Entity extraction complete: 5 entities found
+2025-01-15T10:00:32Z DEBUG Fact extraction complete: 3 facts found
+2025-01-15T10:00:33Z INFO  Contradiction check: 0 contradictions detected
+2025-01-15T10:00:34Z INFO  Ingest complete: chunk-abc123 (145ms)
+2025-01-15T10:00:40Z WARN  Slow query detected: 850ms for hybrid search
+2025-01-15T10:00:45Z ERROR Pod OOMKilled: poimen-worker-abc123 (memory limit exceeded)
+2025-01-15T10:00:50Z INFO  Pod restarted: poimen-worker-abc123 (restart count: 1)