diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..c18dd8d
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1 @@
+__pycache__/
diff --git a/sentinel/.trace-tests-config.yml b/sentinel/.trace-tests-config.yml
index e69de29..6921a62 100644
--- a/sentinel/.trace-tests-config.yml
+++ b/sentinel/.trace-tests-config.yml
@@ -0,0 +1,5 @@
+trace_tests:
+  enabled: true
+  test_data_dir: ./test_data
+  output_dir: ./test_reports
+  max_traces_per_run: 100
\ No newline at end of file
diff --git a/sentinel/docker-compose.yml b/sentinel/docker-compose.yml
index e69de29..0accd48 100644
--- a/sentinel/docker-compose.yml
+++ b/sentinel/docker-compose.yml
@@ -0,0 +1,13 @@
+version: '3.8'
+
+services:
+  sentinel:
+    build: .
+    ports:
+      - "8001:8001"
+    environment:
+      - QUARANTINE_THRESHOLD=0.7
+      - TRACE_PRIVATE_KEY_PEM=optional
+    volumes:
+      - ./data:/app/data
+    restart: unless-stopped
\ No newline at end of file
diff --git a/sentinel/integration.yaml b/sentinel/integration.yaml
index 81d8304..a38f4d5 100644
--- a/sentinel/integration.yaml
+++ b/sentinel/integration.yaml
@@ -7,7 +7,9 @@ metadata:
   category: governance
   maintainer:
     name: Akhilesh Warik
-    email: akhilesh.warik@example.com
+
+    email: warik.akhilesh@gmail.com
+
     github: a1k7
   license: MIT
   version: 2.0.0
@@ -28,4 +30,4 @@ spec:
   evidence:
     - type: manual
       description: |
-        Agent Sentinel produces a risk score, detection list, quarantine action, and collusion patterns.
\ No newline at end of file
+        Agent Sentinel produces a risk score, detection list, quarantine action, and collusion patterns.
diff --git a/sentinel/src/detectors/collusion_detector.py b/sentinel/src/detectors/collusion_detector.py
index 46f72a4..9b374fc 100644
--- a/sentinel/src/detectors/collusion_detector.py
+++ b/sentinel/src/detectors/collusion_detector.py
@@ -64,4 +64,8 @@ def _risk_level(self, score: float) -> RiskLevel:
         if score < 0.3: return RiskLevel.LOW
         if score < 0.6: return RiskLevel.MEDIUM
         if score < 0.8: return RiskLevel.HIGH
-        return RiskLevel.CRITICAL
\ No newline at end of file
+        return RiskLevel.CRITICAL
+if __name__ == "__main__":
+    # Quick test
+    from ..models import DetectionResult, RiskLevel
+    print("CollusionDetector loaded successfully.")
\ No newline at end of file
diff --git a/sentinel/src/trace_claim_generator.py b/sentinel/src/trace_claim_generator.py
index 3e87588..393a073 100644
--- a/sentinel/src/trace_claim_generator.py
+++ b/sentinel/src/trace_claim_generator.py
@@ -4,12 +4,9 @@
 import base64
 from datetime import datetime
 from typing import Dict, Any
+
 from src.models import DetectionResult, TraceClaim, DetectionType
 
-import time
-import hashlib
-from datetime import datetime
-from src.models import DetectionResult, TraceClaim
 
 def generate_trace_claim(agent_id: str, detection: DetectionResult, decision: str = "ADMIT") -> TraceClaim:
     claim_id = f"sentinel-{int(time.time())}-{hashlib.md5(f'{agent_id}{detection.detection_type}'.encode()).hexdigest()[:8]}"
@@ -40,6 +37,7 @@ def generate_trace_claim(agent_id: str, detection: DetectionResult, decision: st
         decision=decision
     )
 
+
 def export_trace_claim(claim: TraceClaim, format: str = "json") -> str:
     """Export the trace claim as JSON or JWT format."""
     if format == "json":
@@ -47,4 +45,4 @@ def export_trace_claim(claim: TraceClaim, format: str = "json") -> str:
     elif format == "jwt":
         return claim.jwt or "JWT not generated (set TRACE_PRIVATE_KEY_PEM)"
     else:
-        raise ValueError(f"Unsupported export format: {format}")
\ No newline at end of file
+        raise ValueError(f"Unsupported export format: {format}")
diff --git a/sentinel/tests/test_detectors.py b/sentinel/tests/test_detectors.py
index e69de29..caad6ff 100644
--- a/sentinel/tests/test_detectors.py
+++ b/sentinel/tests/test_detectors.py
@@ -0,0 +1,27 @@
+import pytest
+from src.detectors import DelegationEscalationDetector
+from src.models import DetectionResult, DetectionType, RiskLevel
+
+def test_delegation_escalation_detector():
+    detector = DelegationEscalationDetector()
+    # Simulate a trace with delegation chain length > 2
+    trace = {
+        "agent_id": "test-agent",
+        "delegation_chain": ["root", "admin", "superadmin"],  # length 3 > threshold 2
+        "policy_version": "v1"
+    }
+    result = detector.detect(trace)
+    assert result is not None
+    assert result.detection_type == DetectionType.DELEGATION_ESCALATION
+    assert result.risk_score > 0.5
+    assert result.risk_level in (RiskLevel.MEDIUM, RiskLevel.HIGH)
+
+def test_delegation_escalation_detector_clean():
+    detector = DelegationEscalationDetector()
+    trace = {
+        "agent_id": "test-agent",
+        "delegation_chain": ["root", "admin"],  # length 2 allowed
+        "policy_version": "v1"
+    }
+    result = detector.detect(trace)
+    assert result is None or result.risk_score < 0.5
\ No newline at end of file
diff --git a/sentinel/tests/test_integration.py b/sentinel/tests/test_integration.py
index e69de29..e2c9de7 100644
--- a/sentinel/tests/test_integration.py
+++ b/sentinel/tests/test_integration.py
@@ -0,0 +1,20 @@
+import pytest
+import json
+from src.cli import main as cli_main
+from src.server import app
+from fastapi.testclient import TestClient
+
+client = TestClient(app)
+
+def test_health_endpoint():
+    response = client.get("/health")
+    assert response.status_code == 200
+    assert response.json() == {"status": "ok"}
+
+def test_cli_import():
+    # Smoke test: ensure CLI can be imported without errors
+    try:
+        from src.cli import main
+        assert callable(main)
+    except ImportError:
+        pytest.fail("CLI module not importable")
\ No newline at end of file