=== Evaluation Results ===
Results file: mcpchecker-openai-agent-kubernetes-test-out.json
=== Evaluation Summary ===

  ✓ ahoy (assertions: 3/3)
  ✓ create-canary-deployment (assertions: 3/3)
  ✓ create-network-policy (assertions: 3/3)
  ✓ create-nginx-pod (assertions: 3/3)
  ✓ create-pod-mount-configmaps (assertions: 3/3)
  ✓ create-pod-resources-limits (assertions: 3/3)
  ✓ create-simple-rbac (assertions: 3/3)
  ✓ debug-app-logs (assertions: 3/3)
  ✓ deployment-traffic-switch (assertions: 3/3)
  ✓ filter-events-by-type (assertions: 3/3)
  ✓ filter-namespace-by-name (assertions: 3/3)
  ✓ fix-crashloop (assertions: 3/3)
  ✓ fix-image-pull (assertions: 3/3)
  ✓ fix-pending-pod (assertions: 3/3)
  ~ fix-probes (assertions: 2/3)
      - MaxToolCalls: Too many tool calls: expected <= 20, got 26
  ✓ fix-rbac-wrong-resource (assertions: 3/3)
  ~ fix-service-routing (assertions: 2/3)
      - MaxToolCalls: Too many tool calls: expected <= 20, got 33
  ✓ fix-service-with-no-endpoints (assertions: 3/3)
  ✓ horizontal-pod-autoscaler (assertions: 3/3)
  ✓ list-images-for-pods (assertions: 3/3)
  ✓ multi-container-pod-communication (assertions: 3/3)
  ✓ nodes-top (assertions: 3/3)
  ✓ resize-pvc (assertions: 3/3)
  ✓ rolling-update-deployment (assertions: 3/3)
  ✓ scale-deployment (assertions: 3/3)
  ✓ scale-down-deployment (assertions: 3/3)
  ~ setup-dev-cluster (assertions: 2/3)
      - MaxToolCalls: Too many tool calls: expected <= 20, got 44
  ✓ ssa-field-preservation (assertions: 3/3)
  ✓ statefulset-lifecycle (assertions: 3/3)

Tasks:      29/29 passed (100.00%)
Assertions: 84/87 passed (96.55%)
Tokens:     ~1612835 (estimate - excludes system prompt & cache)
MCP schemas: ~96686 (included in token total)
Agent used tokens:
  Input:  269997 tokens
  Output: 143990 tokens
Judge used tokens:
  Input:  7663 tokens
  Output: 1381 tokens
=== Threshold Verification ===

Task Pass Rate:      100.00% >= 80.00% ✓
Assertion Pass Rate: 96.55% >= 80.00% ✓

Result: PASSED