diff --git a/agent-framework/prometheus_swarm/utils/duplicate_evidence.py b/agent-framework/prometheus_swarm/utils/duplicate_evidence.py
new file mode 100644
index 00000000..87986aae
--- /dev/null
+++ b/agent-framework/prometheus_swarm/utils/duplicate_evidence.py
@@ -0,0 +1,111 @@
+import logging
+from typing import List, Dict, Any, Optional
+
+class DuplicateEvidenceError(Exception):
+    """Custom exception for duplicate evidence scenarios."""
+    pass
+
+def validate_unique_evidence(evidence_list: List[Dict[Any, Any]], 
+                              unique_key: str = 'id') -> None:
+    """
+    Validate that evidence entries are unique based on a specified key.
+
+    Args:
+        evidence_list (List[Dict[Any, Any]]): List of evidence dictionaries
+        unique_key (str, optional): Key used to determine uniqueness. Defaults to 'id'.
+
+    Raises:
+        DuplicateEvidenceError: If duplicate evidence is detected
+    """
+    logger = logging.getLogger(__name__)
+    
+    # Check for duplicates
+    seen_keys = set()
+    duplicates = []
+
+    for item in evidence_list:
+        if unique_key not in item:
+            logger.warning(f"Evidence item missing unique key '{unique_key}': {item}")
+            continue
+
+        current_key = item[unique_key]
+        
+        if current_key in seen_keys:
+            duplicates.append(current_key)
+            logger.error(f"Duplicate evidence found with {unique_key}: {current_key}")
+        
+        seen_keys.add(current_key)
+
+    if duplicates:
+        raise DuplicateEvidenceError(
+            f"Duplicate evidence detected for {unique_key}s: {duplicates}"
+        )
+
+def log_evidence_summary(evidence_list: List[Dict[Any, Any]], 
+                          log_level: str = 'INFO') -> None:
+    """
+    Log a summary of evidence entries with configurable log level.
+
+    Args:
+        evidence_list (List[Dict[Any, Any]]): List of evidence dictionaries
+        log_level (str, optional): Logging level. Defaults to 'INFO'.
+    """
+    logger = logging.getLogger(__name__)
+    log_method = getattr(logger, log_level.lower(), logger.info)
+
+    log_method(f"Total evidence entries: {len(evidence_list)}")
+    log_method(f"Evidence keys: {list(evidence_list[0].keys()) if evidence_list else 'N/A'}")
+
+def filter_duplicates(evidence_list: List[Dict[Any, Any]], 
+                      unique_key: str = 'id', 
+                      keep: str = 'first') -> List[Dict[Any, Any]]:
+    """
+    Filter out duplicate evidence entries while preserving desired entries.
+
+    Args:
+        evidence_list (List[Dict[Any, Any]]): List of evidence dictionaries
+        unique_key (str, optional): Key used to determine uniqueness. Defaults to 'id'.
+        keep (str, optional): Strategy for keeping duplicates. 
+                               'first' keeps first occurrence, 'last' keeps last. 
+                               Defaults to 'first'.
+
+    Returns:
+        List[Dict[Any, Any]]: Filtered list of evidence without duplicates
+    """
+    logger = logging.getLogger(__name__)
+    
+    if keep not in ['first', 'last']:
+        raise ValueError("'keep' must be either 'first' or 'last'")
+
+    seen_keys = set()
+    filtered_evidence = []
+    non_unique_items = []
+
+    # First pass: handle total list
+    if keep == 'first':
+        for item in evidence_list:
+            if unique_key not in item:
+                non_unique_items.append(item)
+                continue
+
+            current_key = item[unique_key]
+            
+            if current_key not in seen_keys:
+                filtered_evidence.append(item)
+                seen_keys.add(current_key)
+    else:  # keep == 'last'
+        for item in reversed(evidence_list):
+            if unique_key not in item:
+                non_unique_items.insert(0, item)
+                continue
+
+            current_key = item[unique_key]
+            
+            if current_key not in seen_keys:
+                filtered_evidence.insert(0, item)
+                seen_keys.add(current_key)
+
+    # Handle case with non-unique entries
+    filtered_evidence.extend(non_unique_items)
+
+    return filtered_evidence
\ No newline at end of file
diff --git a/agent-framework/tests/unit/test_duplicate_evidence.py b/agent-framework/tests/unit/test_duplicate_evidence.py
new file mode 100644
index 00000000..dc3a1294
--- /dev/null
+++ b/agent-framework/tests/unit/test_duplicate_evidence.py
@@ -0,0 +1,114 @@
+import pytest
+import logging
+from typing import List, Dict
+from prometheus_swarm.utils.duplicate_evidence import (
+    validate_unique_evidence, 
+    DuplicateEvidenceError, 
+    log_evidence_summary,
+    filter_duplicates
+)
+
+def test_validate_unique_evidence_no_duplicates():
+    """Test validation of unique evidence passes."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'id': 2, 'data': 'second'},
+        {'id': 3, 'data': 'third'}
+    ]
+    
+    try:
+        validate_unique_evidence(evidence)
+    except DuplicateEvidenceError:
+        pytest.fail("Unexpected DuplicateEvidenceError raised")
+
+def test_validate_unique_evidence_with_duplicates():
+    """Test validation raises error for duplicate evidence."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'id': 2, 'data': 'second'},
+        {'id': 1, 'data': 'duplicate'}
+    ]
+    
+    with pytest.raises(DuplicateEvidenceError, match="Duplicate evidence detected"):
+        validate_unique_evidence(evidence)
+
+def test_validate_unique_evidence_missing_key():
+    """Test behavior with evidence missing unique key."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'data': 'no id'},
+        {'id': 2, 'data': 'second'}
+    ]
+    
+    # Should not raise an error, just log a warning
+    validate_unique_evidence(evidence)
+
+def test_log_evidence_summary(caplog):
+    """Test logging of evidence summary."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'id': 2, 'data': 'second'}
+    ]
+    
+    with caplog.at_level(logging.INFO):
+        log_evidence_summary(evidence)
+        
+    assert "Total evidence entries: 2" in caplog.text
+
+def test_filter_duplicates_first_occurrence():
+    """Test filtering duplicates, keeping first occurrence."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'id': 2, 'data': 'second'},
+        {'id': 1, 'data': 'duplicate'}
+    ]
+    
+    filtered = filter_duplicates(evidence)
+    
+    assert len(filtered) == 2
+    assert filtered == [
+        {'id': 1, 'data': 'first'},
+        {'id': 2, 'data': 'second'}
+    ]
+
+def test_filter_duplicates_last_occurrence():
+    """Test filtering duplicates, keeping last occurrence."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'id': 2, 'data': 'second'},
+        {'id': 1, 'data': 'duplicate'}
+    ]
+    
+    filtered = filter_duplicates(evidence, keep='last')
+    
+    assert len(filtered) == 2
+    assert filtered == [
+        {'id': 2, 'data': 'second'},
+        {'id': 1, 'data': 'duplicate'}
+    ]
+
+def test_filter_duplicates_invalid_keep_strategy():
+    """Test that an invalid keep strategy raises an error."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'id': 2, 'data': 'second'}
+    ]
+    
+    with pytest.raises(ValueError, match="'keep' must be either 'first' or 'last'"):
+        filter_duplicates(evidence, keep='invalid')
+
+def test_filter_duplicates_missing_key():
+    """Test filtering duplicates with entries missing unique key."""
+    evidence = [
+        {'id': 1, 'data': 'first'},
+        {'data': 'no id'},
+        {'id': 1, 'data': 'duplicate'}
+    ]
+    
+    filtered = filter_duplicates(evidence)
+    
+    assert len(filtered) == 2
+    assert filtered == [
+        {'id': 1, 'data': 'first'},
+        {'data': 'no id'}
+    ]
\ No newline at end of file