[fix]fix token auth (#69)

* fix tocken auth

* Further fixes to the token overwriting issue and restoration of hot reloading of tokens.json.

doris_mcp_server/main.py

@@ -432,9 +432,9 @@ class DorisServer:
             await self.security_manager.initialize()
             self.logger.info("Security manager initialization completed")
             
-            # Ensure connection manager is initialized
-            await self.connection_manager.initialize()
-            self.logger.info("Connection manager initialization completed")
+            # For stdio mode, we must establish a working database connection
+            # Use the dedicated stdio mode initialization method
+            await self.connection_manager.initialize_for_stdio_mode()
 
             # Start stdio server - using compatible import approach
             try:
@@ -502,8 +502,12 @@ class DorisServer:
             await self.security_manager.initialize()
             self.logger.info("Security manager initialization completed")
             
-            # Ensure connection manager is initialized
-            await self.connection_manager.initialize()
+            # For HTTP mode, try to initialize global connection pool with graceful degradation
+            global_pool_created = await self.connection_manager.initialize_for_http_mode()
+            if global_pool_created:
+                self.logger.info("Global database connection pool available for HTTP mode")
+            else:
+                self.logger.info("HTTP mode running without global database pool, will use token-bound configurations")
 
             # Use Starlette and StreamableHTTPSessionManager according to official example
             import uvicorn
@@ -638,6 +642,16 @@ class DorisServer:
                                 # Store auth context in scope for potential use by tools/resources
                                 scope["auth_context"] = auth_context
 
+                                # FIX for Issue #62 Bug 1: Set auth_context in context variable
+                                # This allows tools to access token information for token-bound database configuration
+                                # CRITICAL: Use the global ContextVar from security.py to ensure same instance is used everywhere
+                                try:
+                                    from .utils.security import mcp_auth_context_var
+                                    mcp_auth_context_var.set(auth_context)
+                                    self.logger.debug(f"Set auth_context in context variable with token: {bool(hasattr(auth_context, 'token') and auth_context.token)}")
+                                except Exception as ctx_error:
+                                    self.logger.warning(f"Failed to set auth_context in context variable: {ctx_error}")
+
                             except Exception as auth_error:
                                 self.logger.error(f"MCP authentication failed: {auth_error}")
                                 # Return 401 Unauthorized

doris_mcp_server/tools/prompts_manager.py

@@ -31,6 +31,7 @@ from mcp.types import (
 )
 
 from ..utils.db import DorisConnectionManager
+from ..utils.sql_security_utils import get_auth_context
 
 
 class PromptTemplate:
@@ -422,7 +423,8 @@ Please generate accurate and efficient SQL queries based on the above requiremen
             AND table_type = 'BASE TABLE'
             """
 
-            db_result = await connection.execute(db_info_sql)
+            auth_context = get_auth_context()
+            db_result = await connection.execute(db_info_sql, auth_context=auth_context)
             db_info = db_result.data[0] if db_result.data else {}
 
             # Get main table list
@@ -438,7 +440,7 @@ Please generate accurate and efficient SQL queries based on the above requiremen
             LIMIT 10
             """
 
-            tables_result = await connection.execute(tables_sql)
+            tables_result = await connection.execute(tables_sql, auth_context=auth_context)
 
             context = f"""Current database statistics:
 - Total number of tables: {db_info.get("table_count", 0)}

doris_mcp_server/tools/resources_manager.py

@@ -26,6 +26,7 @@ from typing import Any
 from mcp.types import Resource
 
 from ..utils.db import DorisConnectionManager
+from ..utils.sql_security_utils import get_auth_context
 
 
 class TableMetadata:
@@ -169,7 +170,8 @@ class DorisResourcesManager:
         ORDER BY table_name
         """
 
-        result = await connection.execute(tables_query)
+        auth_context = get_auth_context()
+        result = await connection.execute(tables_query, auth_context=auth_context)
         tables = []
 
         for row in result.data:
@@ -204,7 +206,8 @@ class DorisResourcesManager:
         ORDER BY ordinal_position
         """
 
-        result = await connection.execute(columns_query, (table_name,))
+        auth_context = get_auth_context()
+        result = await connection.execute(columns_query, params=(table_name,), auth_context=auth_context)
         return [dict(row) for row in result.data]
 
     async def _get_view_metadata(self) -> list[ViewMetadata]:
@@ -226,7 +229,8 @@ class DorisResourcesManager:
         ORDER BY table_name
         """
 
-        result = await connection.execute(views_query)
+        auth_context = get_auth_context()
+        result = await connection.execute(views_query, auth_context=auth_context)
         views = []
 
         for row in result.data:
@@ -257,7 +261,8 @@ class DorisResourcesManager:
         AND table_name = %s
         """
 
-        table_result = await connection.execute(table_info_query, (table_name,))
+        auth_context = get_auth_context()
+        table_result = await connection.execute(table_info_query, params=(table_name,), auth_context=auth_context)
         if not table_result.data:
             raise ValueError(f"Table {table_name} does not exist")
 
@@ -295,7 +300,8 @@ class DorisResourcesManager:
         ORDER BY index_name, seq_in_index
         """
 
-        result = await connection.execute(indexes_query, (table_name,))
+        auth_context = get_auth_context()
+        result = await connection.execute(indexes_query, params=(table_name,), auth_context=auth_context)
         return [dict(row) for row in result.data]
 
     async def _get_view_definition(self, view_name: str) -> str:
@@ -312,7 +318,8 @@ class DorisResourcesManager:
         AND table_name = %s
         """
 
-        result = await connection.execute(view_query, (view_name,))
+        auth_context = get_auth_context()
+        result = await connection.execute(view_query, params=(view_name,), auth_context=auth_context)
         if not result.data:
             raise ValueError(f"View {view_name} does not exist")
 
@@ -340,7 +347,8 @@ class DorisResourcesManager:
         AND table_type = 'BASE TABLE'
         """
 
-        table_result = await connection.execute(table_stats_query)
+        auth_context = get_auth_context()
+        table_result = await connection.execute(table_stats_query, auth_context=auth_context)
         table_stats = table_result.data[0] if table_result.data else {}
 
         # Get view statistics
@@ -350,7 +358,7 @@ class DorisResourcesManager:
         WHERE table_schema = DATABASE()
         """
 
-        view_result = await connection.execute(view_stats_query)
+        view_result = await connection.execute(view_stats_query, auth_context=auth_context)
         view_stats = view_result.data[0] if view_result.data else {}
 
         stats_info = {

doris_mcp_server/utils/adbc_query_tools.py

@@ -28,6 +28,7 @@ from typing import Any, Dict, List, Optional
 
 from ..utils.logger import get_logger
 from ..utils.db import DorisConnectionManager
+from ..utils.sql_security_utils import get_auth_context
 
 logger = get_logger(__name__)
 
@@ -277,7 +278,8 @@ class DorisADBCQueryTools:
             # Get BE nodes via SHOW BACKENDS
             logger.info("No BE hosts configured, getting BE node information via SHOW BACKENDS")
             connection = await self.connection_manager.get_connection("query")
-            result = await connection.execute("SHOW BACKENDS")
+            auth_context = get_auth_context()
+            result = await connection.execute("SHOW BACKENDS", auth_context=auth_context)
             
             be_hosts = []
             for row in result.data:
@@ -383,6 +385,20 @@ class DorisADBCQueryTools:
                     "error_type": "no_connection"
                 }
             
+            # SECURITY FIX: Perform SQL security validation before executing
+            auth_context = get_auth_context()
+            if self.connection_manager.security_manager:
+                # Always perform security validation, even without auth_context
+                # Use a default context for basic SQL security checks
+                validation_result = await self.connection_manager.security_manager.validate_sql_security(sql, auth_context)
+                if not validation_result.is_valid:
+                    return {
+                        "success": False,
+                        "error": f"SQL security validation failed: {validation_result.error_message}",
+                        "error_type": "security_violation",
+                        "risk_level": validation_result.risk_level
+                    }
+            
             cursor = self.adbc_client.cursor()
             start_time = time.time()
             

doris_mcp_server/utils/analysis_tools.py

@@ -29,6 +29,13 @@ from pathlib import Path
 
 from .db import DorisConnectionManager
 from .logger import get_logger
+from .sql_security_utils import (
+    SQLSecurityError,
+    validate_identifier,
+    quote_identifier,
+    build_table_reference,
+    get_auth_context
+)
 
 logger = get_logger(__name__)
 
@@ -46,10 +53,17 @@ class TableAnalyzer:
         sample_size: int = 10
     ) -> Dict[str, Any]:
         """Get table summary information"""
+        # SECURITY FIX: Validate table_name and get auth_context
+        try:
+            validate_identifier(table_name, "table name")
+        except SQLSecurityError as e:
+            raise ValueError(f"Invalid table name: {e}")
+        
+        auth_context = get_auth_context()
         connection = await self.connection_manager.get_connection("query")
         
-        # Get table basic information
-        table_info_sql = f"""
+        # Get table basic information using parameterized query
+        table_info_sql = """
         SELECT 
             table_name,
             table_comment,
@@ -58,17 +72,17 @@ class TableAnalyzer:
             engine
         FROM information_schema.tables 
         WHERE table_schema = DATABASE()
-        AND table_name = '{table_name}'
+        AND table_name = %s
         """
         
-        table_info_result = await connection.execute(table_info_sql)
+        table_info_result = await connection.execute(table_info_sql, params=(table_name,), auth_context=auth_context)
         if not table_info_result.data:
             raise ValueError(f"Table {table_name} does not exist")
         
         table_info = table_info_result.data[0]
         
-        # Get column information
-        columns_sql = f"""
+        # Get column information using parameterized query
+        columns_sql = """
         SELECT 
             column_name,
             data_type,
@@ -76,11 +90,11 @@ class TableAnalyzer:
             column_comment
         FROM information_schema.columns 
         WHERE table_schema = DATABASE()
-        AND table_name = '{table_name}'
+        AND table_name = %s
         ORDER BY ordinal_position
         """
         
-        columns_result = await connection.execute(columns_sql)
+        columns_result = await connection.execute(columns_sql, params=(table_name,), auth_context=auth_context)
         
         summary = {
             "table_name": table_info["table_name"],
@@ -92,10 +106,11 @@ class TableAnalyzer:
             "columns": columns_result.data,
         }
         
-        # Get sample data
+        # Get sample data using quoted identifier
         if include_sample and sample_size > 0:
-            sample_sql = f"SELECT * FROM {table_name} LIMIT {sample_size}"
-            sample_result = await connection.execute(sample_sql)
+            quoted_table = quote_identifier(table_name, "table name")
+            sample_sql = f"SELECT * FROM {quoted_table} LIMIT {sample_size}"
+            sample_result = await connection.execute(sample_sql, auth_context=auth_context)
             summary["sample_data"] = sample_result.data
         
         return summary
@@ -120,7 +135,8 @@ class TableAnalyzer:
             FROM {table_name}
             """
             
-            basic_result = await connection.execute(basic_stats_sql)
+            auth_context = get_auth_context()
+            basic_result = await connection.execute(basic_stats_sql, auth_context=auth_context)
             if not basic_result.data:
                 return {
                     "success": False,
@@ -144,7 +160,7 @@ class TableAnalyzer:
                 LIMIT 20
                 """
                 
-                distribution_result = await connection.execute(distribution_sql)
+                distribution_result = await connection.execute(distribution_sql, auth_context=auth_context)
                 analysis["value_distribution"] = distribution_result.data
             
             if analysis_type == "detailed":
@@ -159,7 +175,7 @@ class TableAnalyzer:
                     WHERE {column_name} IS NOT NULL
                     """
                     
-                    numeric_result = await connection.execute(numeric_stats_sql)
+                    numeric_result = await connection.execute(numeric_stats_sql, auth_context=auth_context)
                     if numeric_result.data:
                         analysis.update(numeric_result.data[0])
                 except Exception:
@@ -196,7 +212,8 @@ class TableAnalyzer:
         AND table_name = '{table_name}'
         """
         
-        table_result = await connection.execute(table_info_sql)
+        auth_context = get_auth_context()
+        table_result = await connection.execute(table_info_sql, auth_context=auth_context)
         if not table_result.data:
             raise ValueError(f"Table {table_name} does not exist")
         
@@ -211,7 +228,7 @@ class TableAnalyzer:
         AND table_name != %s
         """
         
-        all_tables_result = await connection.execute(all_tables_sql, (table_name,))
+        all_tables_result = await connection.execute(all_tables_sql, params=(table_name,), auth_context=auth_context)
         
         return {
             "center_table": table_result.data[0],
@@ -291,7 +308,8 @@ class PerformanceMonitor:
             LIMIT 20
             """
             
-            tables_result = await connection.execute(tables_sql)
+            auth_context = get_auth_context()
+            tables_result = await connection.execute(tables_sql, auth_context=auth_context)
             stats = {
                 "metric_type": "tables",
                 "time_range": time_range,
@@ -379,9 +397,23 @@ class SQLAnalyzer:
             
             logger.info(f"Generating SQL explain for query ID: {query_id}")
             
+            # 🔧 FIX: Get auth_context for token-bound database configuration
+            auth_context = None
+            try:
+                from .security import mcp_auth_context_var
+                auth_context = mcp_auth_context_var.get()
+            except Exception:
+                pass
+            
             # Switch database if specified
+            # SECURITY FIX: Validate and quote db_name
             if db_name:
-                await self.connection_manager.execute_query("explain_session", f"USE {db_name}")
+                try:
+                    validate_identifier(db_name, "database name")
+                except SQLSecurityError as e:
+                    return {"success": False, "error": f"Invalid database name: {e}"}
+                safe_db = quote_identifier(db_name, "database name")
+                await self.connection_manager.execute_query("explain_session", f"USE {safe_db}", None, auth_context)
             
             # Construct EXPLAIN query
             explain_type = "EXPLAIN VERBOSE" if verbose else "EXPLAIN"
@@ -390,7 +422,7 @@ class SQLAnalyzer:
             logger.info(f"Executing explain query: {explain_sql}")
             
             # Execute explain query
-            result = await self.connection_manager.execute_query("explain_session", explain_sql)
+            result = await self.connection_manager.execute_query("explain_session", explain_sql, None, auth_context)
             
             # Format explain output
             explain_content = []
@@ -515,24 +547,36 @@ class SQLAnalyzer:
             
             try:
                 # Switch to specified database/catalog if provided
+                # SECURITY FIX: Validate identifiers before using in SQL
                 if catalog_name:
-                    await connection.execute(f"SWITCH `{catalog_name}`")
+                    try:
+                        validate_identifier(catalog_name, "catalog name")
+                    except SQLSecurityError as e:
+                        return {"success": False, "error": f"Invalid catalog name: {e}"}
+                    safe_catalog = quote_identifier(catalog_name, "catalog name")
+                    auth_context = get_auth_context()
+                    await connection.execute(f"SWITCH {safe_catalog}", auth_context=auth_context)
                 if db_name:
-                    await connection.execute(f"USE `{db_name}`")
+                    try:
+                        validate_identifier(db_name, "database name")
+                    except SQLSecurityError as e:
+                        return {"success": False, "error": f"Invalid database name: {e}"}
+                    safe_db = quote_identifier(db_name, "database name")
+                    await connection.execute(f"USE {safe_db}", auth_context=auth_context)
                 
                 # Set trace ID for the session using session variable
                 # According to official docs: set session_context="trace_id:your_trace_id"
-                await connection.execute(f'set session_context="trace_id:{trace_id}"')
+                await connection.execute(f'set session_context="trace_id:{trace_id}"', auth_context=auth_context)
                 logger.info(f"Set trace ID: {trace_id}")
 
                 # Enable profile
-                await connection.execute(f'set enable_profile=true')
+                await connection.execute(f'set enable_profile=true', auth_context=auth_context)
                 logger.info(f"Enabled profile")
                 
                 # Execute the SQL statement
                 logger.info(f"Executing SQL with trace ID: {sql}")
                 start_time = time.time()
-                sql_result = await connection.execute(sql)
+                sql_result = await connection.execute(sql, auth_context=auth_context)
                 execution_time = time.time() - start_time
                 logger.info(f"SQL execution completed in {execution_time:.3f}s")
                 

doris_mcp_server/utils/data_exploration_tools.py

@@ -26,6 +26,13 @@ from typing import Any, Dict, List, Optional, Union
 
 from .db import DorisConnectionManager
 from .logger import get_logger
+from .sql_security_utils import (
+    SQLSecurityError,
+    validate_identifier,
+    quote_identifier,
+    build_table_reference,
+    get_auth_context
+)
 
 logger = get_logger(__name__)
 
@@ -43,24 +50,30 @@ class DataExplorationTools:
     
     def _build_full_table_name(self, table_name: str, catalog_name: Optional[str], db_name: Optional[str]) -> str:
         """Build full table name with catalog and database using three-part naming convention"""
-        # Default catalog for internal tables
+        # SECURITY FIX: Use build_table_reference for safe identifier handling
         effective_catalog = catalog_name if catalog_name else "internal"
         
         if db_name:
-            return f"{effective_catalog}.{db_name}.{table_name}"
+            return build_table_reference(table_name, db_name, effective_catalog)
         else:
-            # If no db_name provided, need to determine the current database
-            return f"{effective_catalog}.{table_name}"
+            return build_table_reference(table_name, catalog_name=effective_catalog)
     
     async def _get_table_basic_info(self, connection, table_name: str) -> Optional[Dict]:
         """Get basic table information including row count"""
         try:
+            # SECURITY FIX: Get auth_context for security validation
+            # table_name should already be validated by _build_full_table_name
+            auth_context = get_auth_context()
+            
             count_sql = f"SELECT COUNT(*) as row_count FROM {table_name}"
-            result = await connection.execute(count_sql)
+            result = await connection.execute(count_sql, auth_context=auth_context)
             
             if result.data:
                 return {"row_count": result.data[0]["row_count"]}
             return None
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed for table {table_name}: {str(e)}")
+            return {"row_count": 0}
         except Exception as e:
             logger.warning(f"Failed to get basic info for table {table_name}: {str(e)}")
             return {"row_count": 0}
@@ -68,10 +81,24 @@ class DataExplorationTools:
     async def _get_table_columns_info(self, connection, table_name: str, catalog_name: Optional[str], db_name: Optional[str]) -> List[Dict]:
         """Get detailed column information"""
         try:
-            where_conditions = [f"table_name = '{table_name}'"]
+            # SECURITY FIX: Validate identifiers and use parameterized query
+            auth_context = get_auth_context()
+            
+            try:
+                validate_identifier(table_name, "table name")
+                if db_name:
+                    validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            # Build parameterized query
+            params = [table_name]
+            where_conditions = ["table_name = %s"]
             
             if db_name:
-                where_conditions.append(f"table_schema = '{db_name}'")
+                where_conditions.append("table_schema = %s")
+                params.append(db_name)
             else:
                 where_conditions.append("table_schema = DATABASE()")
             
@@ -87,9 +114,12 @@ class DataExplorationTools:
             ORDER BY ordinal_position
             """
             
-            result = await connection.execute(columns_sql)
+            result = await connection.execute(columns_sql, params=tuple(params), auth_context=auth_context)
             return result.data if result.data else []
             
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed: {str(e)}")
+            return []
         except Exception as e:
             logger.warning(f"Failed to get columns info for table {table_name}: {str(e)}")
             return []
@@ -177,7 +207,8 @@ class DataExplorationTools:
                 WHERE {col_name} IS NOT NULL
                 """
                 
-                stats_result = await connection.execute(stats_sql)
+                auth_context = get_auth_context()
+                stats_result = await connection.execute(stats_sql, auth_context=auth_context)
                 
                 if stats_result.data and stats_result.data[0]["count"] > 0:
                     stats = stats_result.data[0]
@@ -229,7 +260,8 @@ class DataExplorationTools:
             WHERE {col_name} IS NOT NULL
             """
             
-            result = await connection.execute(percentile_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(percentile_sql, auth_context=auth_context)
             
             if result.data:
                 data = result.data[0]
@@ -268,7 +300,8 @@ class DataExplorationTools:
             WHERE {col_name} IS NOT NULL
             """
             
-            result = await connection.execute(outlier_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(outlier_sql, auth_context=auth_context)
             
             if result.data:
                 data = result.data[0]
@@ -359,7 +392,8 @@ class DataExplorationTools:
                 {sampling_info.get('sample_query_suffix', '')}
                 """
                 
-                cardinality_result = await connection.execute(cardinality_sql)
+                auth_context = get_auth_context()
+                cardinality_result = await connection.execute(cardinality_sql, auth_context=auth_context)
                 
                 if cardinality_result.data:
                     cardinality_data = cardinality_result.data[0]
@@ -408,7 +442,8 @@ class DataExplorationTools:
             LIMIT 20
             """
             
-            result = await connection.execute(distribution_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(distribution_sql, auth_context=auth_context)
             
             if result.data:
                 distribution = []
@@ -458,7 +493,8 @@ class DataExplorationTools:
                 WHERE {col_name} IS NOT NULL
                 """
                 
-                range_result = await connection.execute(range_sql)
+                auth_context = get_auth_context()
+                range_result = await connection.execute(range_sql, auth_context=auth_context)
                 
                 if range_result.data and range_result.data[0]["non_null_count"] > 0:
                     range_data = range_result.data[0]
@@ -539,7 +575,8 @@ class DataExplorationTools:
             ORDER BY day_of_week
             """
             
-            weekly_result = await connection.execute(weekly_pattern_sql)
+            auth_context = get_auth_context()
+            weekly_result = await connection.execute(weekly_pattern_sql, auth_context=auth_context)
             
             weekly_pattern = []
             if weekly_result.data:
@@ -561,7 +598,7 @@ class DataExplorationTools:
             LIMIT 12
             """
             
-            monthly_result = await connection.execute(monthly_trend_sql)
+            monthly_result = await connection.execute(monthly_trend_sql, auth_context=auth_context)
             monthly_trend = "stable"  # Simplified trend analysis
             
             if monthly_result.data and len(monthly_result.data) > 3:
@@ -646,7 +683,8 @@ class DataExplorationTools:
                 FROM {table_expr}
                 """
                 
-                result = await connection.execute(null_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(null_sql, auth_context=auth_context)
                 if result.data:
                     data = result.data[0]
                     total_count = data["total_count"]

doris_mcp_server/utils/data_governance_tools.py

@@ -26,6 +26,13 @@ from typing import Any, Dict, List, Optional
 
 from .db import DorisConnectionManager
 from .logger import get_logger
+from .sql_security_utils import (
+    SQLSecurityError,
+    validate_identifier,
+    quote_identifier,
+    build_table_reference,
+    get_auth_context
+)
 
 logger = get_logger(__name__)
 
@@ -216,26 +223,34 @@ class DataGovernanceTools:
     # ==================== Private Helper Methods ====================
     
     def _build_full_table_name(self, table_name: str, catalog_name: Optional[str], db_name: Optional[str]) -> str:
-        """Build full table name - use three-level naming convention"""
+        """Build full table name - use three-level naming convention with security validation"""
+        # SECURITY FIX: Use build_table_reference for safe identifier handling
         # Default catalog is internal for internal tables
         effective_catalog = catalog_name if catalog_name else "internal"
         
         if db_name:
-            return f"{effective_catalog}.{db_name}.{table_name}"
+            return build_table_reference(table_name, db_name, effective_catalog)
         else:
             # If db_name is not provided, need to determine current database
-            return f"{effective_catalog}.{table_name}"
+            return build_table_reference(table_name, catalog_name=effective_catalog)
     
     async def _get_table_basic_info(self, connection, table_name: str) -> Optional[Dict]:
         """Get table basic information"""
         try:
+            # SECURITY FIX: Get auth_context for security validation
+            # table_name should already be validated by _build_full_table_name
+            auth_context = get_auth_context()
+            
             # Try to get table row count
             count_sql = f"SELECT COUNT(*) as row_count FROM {table_name}"
-            result = await connection.execute(count_sql)
+            result = await connection.execute(count_sql, auth_context=auth_context)
             
             if result.data:
                 return {"row_count": result.data[0]["row_count"]}
             return None
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed for table {table_name}: {str(e)}")
+            return {"row_count": 0}
         except Exception as e:
             logger.warning(f"Failed to get basic info for table {table_name}: {str(e)}")
             return {"row_count": 0}
@@ -243,11 +258,24 @@ class DataGovernanceTools:
     async def _get_table_columns_info(self, connection, table_name: str, catalog_name: Optional[str], db_name: Optional[str]) -> List[Dict]:
         """Get table column information"""
         try:
-            # Build query conditions
-            where_conditions = [f"table_name = '{table_name}'"]
+            # SECURITY FIX: Validate identifiers and use parameterized query
+            auth_context = get_auth_context()
+            
+            try:
+                validate_identifier(table_name, "table name")
+                if db_name:
+                    validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            # Build parameterized query conditions
+            params = [table_name]
+            where_conditions = ["table_name = %s"]
             
             if db_name:
-                where_conditions.append(f"table_schema = '{db_name}'")
+                where_conditions.append("table_schema = %s")
+                params.append(db_name)
             else:
                 where_conditions.append("table_schema = DATABASE()")
             
@@ -263,30 +291,49 @@ class DataGovernanceTools:
             ORDER BY ordinal_position
             """
             
-            result = await connection.execute(columns_sql)
+            result = await connection.execute(columns_sql, params=tuple(params), auth_context=auth_context)
             return result.data if result.data else []
             
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed: {str(e)}")
+            return []
         except Exception as e:
             logger.warning(f"Failed to get columns info for table {table_name}: {str(e)}")
             return []
     
     async def _analyze_column_completeness(self, connection, table_name: str, columns_info: List[Dict]) -> Dict[str, Any]:
         """Analyze column completeness"""
+        # SECURITY FIX: Get auth_context for security validation
+        auth_context = get_auth_context()
         column_completeness = {}
         
         for column in columns_info:
             column_name = column["column_name"]
             try:
+                # SECURITY FIX: Validate column name before using in SQL
+                try:
+                    validate_identifier(column_name, "column name")
+                except SQLSecurityError as e:
+                    logger.warning(f"Invalid column name rejected: {e}")
+                    column_completeness[column_name] = {
+                        "error": f"Invalid column name: {e}",
+                        "completeness_score": 0.0
+                    }
+                    continue
+                
+                # Use quoted identifier for column name
+                quoted_column = quote_identifier(column_name, "column name")
+                
                 # Calculate null value statistics
                 null_sql = f"""
                 SELECT 
                     COUNT(*) as total_count,
-                    COUNT({column_name}) as non_null_count,
-                    COUNT(*) - COUNT({column_name}) as null_count
+                    COUNT({quoted_column}) as non_null_count,
+                    COUNT(*) - COUNT({quoted_column}) as null_count
                 FROM {table_name}
                 """
                 
-                result = await connection.execute(null_sql)
+                result = await connection.execute(null_sql, auth_context=auth_context)
                 if result.data:
                     stats = result.data[0]
                     total_count = stats["total_count"]
@@ -304,6 +351,12 @@ class DataGovernanceTools:
                         "completeness_score": round(completeness_score, 4)
                     }
                     
+            except SQLSecurityError as e:
+                logger.warning(f"Security validation failed for column {column_name}: {str(e)}")
+                column_completeness[column_name] = {
+                    "error": str(e),
+                    "completeness_score": 0.0
+                }
             except Exception as e:
                 logger.warning(f"Failed to analyze completeness for column {column_name}: {str(e)}")
                 column_completeness[column_name] = {
@@ -333,7 +386,8 @@ class DataGovernanceTools:
                 FROM {table_name}
                 """
                 
-                result = await connection.execute(compliance_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(compliance_sql, auth_context=auth_context)
                 if result.data:
                     stats = result.data[0]
                     pass_count = stats["pass_count"] or 0
@@ -378,7 +432,8 @@ class DataGovernanceTools:
                 ) t
                 """
                 
-                result = await connection.execute(duplicate_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(duplicate_sql, auth_context=auth_context)
                 if result.data and result.data[0]["duplicate_count"] > 0:
                     issues.append({
                         "type": "duplicate_primary_keys",
@@ -456,10 +511,21 @@ class DataGovernanceTools:
     async def _verify_column_exists(self, connection, table_name: str, column_name: str) -> bool:
         """Verify if column exists"""
         try:
-            # Simple verification method: try to query the column
-            verify_sql = f"SELECT {column_name} FROM {table_name} LIMIT 1"
-            await connection.execute(verify_sql)
+            # SECURITY FIX: Validate and quote column name
+            try:
+                validate_identifier(column_name, "column name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid column name rejected: {e}")
+                return False
+            
+            safe_column = quote_identifier(column_name, "column name")
+            # table_name is already safe (from _build_full_table_name)
+            verify_sql = f"SELECT {safe_column} FROM {table_name} LIMIT 1"
+            auth_context = get_auth_context()
+            await connection.execute(verify_sql, auth_context=auth_context)
             return True
+        except SQLSecurityError:
+            return False
         except Exception:
             return False
     
@@ -469,21 +535,34 @@ class DataGovernanceTools:
         source_chain = []
         
         try:
+            # SECURITY FIX: Validate table name and use parameterized-like approach
+            table_name_part = table_name.split('.')[-1]
+            try:
+                validate_identifier(table_name_part, "table name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid table name rejected: {e}")
+                return []
+            
+            # Escape special characters for LIKE pattern
+            safe_pattern = table_name_part.replace('%', r'\%').replace('_', r'\_')
+            like_pattern = f"%{safe_pattern}%"
+            
             # Try to find related INSERT/CREATE TABLE AS SELECT statements from audit logs (one year range)
+            auth_context = get_auth_context()
             audit_sql = """
             SELECT 
                 stmt as sql_statement,
                 `time` as execution_time,
                 `user` as user_name
             FROM internal.__internal_schema.audit_log 
-            WHERE stmt LIKE '%{}%' 
+            WHERE stmt LIKE %s 
                 AND (stmt LIKE '%INSERT%' OR stmt LIKE '%CREATE%' OR stmt LIKE '%SELECT%')
                 AND `time` >= DATE_SUB(NOW(), INTERVAL 1 YEAR)
             ORDER BY `time` DESC 
             LIMIT 50
-            """.format(table_name.split('.')[-1])  # Use the last part of table name
+            """
             
-            result = await connection.execute(audit_sql)
+            result = await connection.execute(audit_sql, params=(like_pattern,), auth_context=auth_context)
             
             if result.data:
                 for i, log_entry in enumerate(result.data[:depth]):
@@ -556,19 +635,33 @@ class DataGovernanceTools:
         downstream_usage = []
         
         try:
+            # SECURITY FIX: Validate inputs and use parameterized-like approach
+            table_name_part = table_name.split('.')[-1]
+            try:
+                validate_identifier(table_name_part, "table name")
+                validate_identifier(column_name, "column name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            # Escape special characters for LIKE pattern
+            safe_table_pattern = f"%{table_name_part.replace('%', r'\\%').replace('_', r'\\_')}%"
+            safe_column_pattern = f"%{column_name.replace('%', r'\\%').replace('_', r'\\_')}%"
+            
             # Find other tables that might use this field (through audit logs, one year range)
+            auth_context = get_auth_context()
             usage_sql = """
             SELECT DISTINCT
                 stmt as sql_statement
             FROM internal.__internal_schema.audit_log 
-            WHERE stmt LIKE '%{}%' 
-                AND stmt LIKE '%{}%'
+            WHERE stmt LIKE %s 
+                AND stmt LIKE %s
                 AND stmt LIKE '%SELECT%'
                 AND `time` >= DATE_SUB(NOW(), INTERVAL 1 YEAR)
             LIMIT 20
-            """.format(table_name.split('.')[-1], column_name)
+            """
             
-            result = await connection.execute(usage_sql)
+            result = await connection.execute(usage_sql, params=(safe_table_pattern, safe_column_pattern), auth_context=auth_context)
             
             if result.data:
                 for entry in result.data:
@@ -634,14 +727,20 @@ class DataGovernanceTools:
     async def _get_all_tables(self, connection, catalog_name: Optional[str], db_name: Optional[str]) -> List[str]:
         """Get list of all tables"""
         try:
-            where_conditions = []
+            auth_context = get_auth_context()
+            params = []
             
+            # SECURITY FIX: Use parameterized query
             if db_name:
-                where_conditions.append(f"table_schema = '{db_name}'")
+                try:
+                    validate_identifier(db_name, "database name")
+                except SQLSecurityError as e:
+                    logger.warning(f"Invalid database name rejected: {e}")
+                    return []
+                where_clause = "table_schema = %s"
+                params.append(db_name)
             else:
-                where_conditions.append("table_schema = DATABASE()")
-            
-            where_clause = " AND ".join(where_conditions) if where_conditions else "1=1"
+                where_clause = "table_schema = DATABASE()"
             
             tables_sql = f"""
             SELECT table_name 
@@ -651,7 +750,7 @@ class DataGovernanceTools:
             ORDER BY table_name
             """
             
-            result = await connection.execute(tables_sql)
+            result = await connection.execute(tables_sql, params=tuple(params) if params else None, auth_context=auth_context)
             return [row["table_name"] for row in result.data] if result.data else []
             
         except Exception as e:
@@ -728,15 +827,23 @@ class DataGovernanceTools:
     async def _get_freshness_from_partition_info(self, connection, table_name: str) -> Optional[Dict]:
         """Get freshness from partition information"""
         try:
-            # Query partition information (if table has partitions)
-            partition_sql = f"""
+            # SECURITY FIX: Validate and use parameterized query
+            table_name_part = table_name.split('.')[-1]
+            try:
+                validate_identifier(table_name_part, "table name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid table name rejected: {e}")
+                return None
+            
+            auth_context = get_auth_context()
+            partition_sql = """
             SELECT MAX(CREATE_TIME) as last_update
             FROM information_schema.partitions 
-            WHERE table_name = '{table_name.split('.')[-1]}'
+            WHERE table_name = %s
                 AND CREATE_TIME IS NOT NULL
             """
             
-            result = await connection.execute(partition_sql)
+            result = await connection.execute(partition_sql, params=(table_name_part,), auth_context=auth_context)
             if result.data and result.data[0]["last_update"]:
                 return {
                     "last_update": result.data[0]["last_update"],
@@ -759,7 +866,8 @@ class DataGovernanceTools:
                 FROM {table_name}
                 """
                 
-                result = await connection.execute(max_time_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(max_time_sql, auth_context=auth_context)
                 if result.data and result.data[0]["last_update"]:
                     return {
                         "last_update": result.data[0]["last_update"],
@@ -773,15 +881,23 @@ class DataGovernanceTools:
     async def _get_freshness_from_table_metadata(self, connection, table_name: str) -> Optional[Dict]:
         """Get freshness from table metadata"""
         try:
-            # Query table's update time
-            metadata_sql = f"""
+            # SECURITY FIX: Validate and use parameterized query
+            table_name_part = table_name.split('.')[-1]
+            try:
+                validate_identifier(table_name_part, "table name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid table name rejected: {e}")
+                return None
+            
+            auth_context = get_auth_context()
+            metadata_sql = """
             SELECT UPDATE_TIME as last_update
             FROM information_schema.tables 
-            WHERE table_name = '{table_name.split('.')[-1]}'
+            WHERE table_name = %s
                 AND UPDATE_TIME IS NOT NULL
             """
             
-            result = await connection.execute(metadata_sql)
+            result = await connection.execute(metadata_sql, params=(table_name_part,), auth_context=auth_context)
             if result.data and result.data[0]["last_update"]:
                 return {
                     "last_update": result.data[0]["last_update"],
@@ -795,10 +911,19 @@ class DataGovernanceTools:
     async def _find_timestamp_columns(self, connection, table_name: str) -> List[str]:
         """Find possible timestamp fields"""
         try:
-            timestamp_sql = f"""
+            # SECURITY FIX: Validate and use parameterized query
+            table_name_part = table_name.split('.')[-1]
+            try:
+                validate_identifier(table_name_part, "table name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid table name rejected: {e}")
+                return []
+            
+            auth_context = get_auth_context()
+            timestamp_sql = """
             SELECT column_name 
             FROM information_schema.columns 
-            WHERE table_name = '{table_name.split('.')[-1]}'
+            WHERE table_name = %s
                 AND (
                     data_type IN ('datetime', 'timestamp', 'date') 
                     OR column_name LIKE '%time%' 
@@ -815,7 +940,7 @@ class DataGovernanceTools:
                 END
             """
             
-            result = await connection.execute(timestamp_sql)
+            result = await connection.execute(timestamp_sql, params=(table_name_part,), auth_context=auth_context)
             return [row["column_name"] for row in result.data] if result.data else []
             
         except Exception:

doris_mcp_server/utils/data_quality_tools.py

@@ -31,6 +31,12 @@ from collections import Counter, defaultdict
 from .db import DorisConnectionManager
 from .logger import get_logger
 from .config import DorisConfig
+from .sql_security_utils import (
+    SQLSecurityError,
+    validate_identifier,
+    build_table_reference,
+    get_auth_context
+)
 
 logger = get_logger(__name__)
 
@@ -299,23 +305,26 @@ class DataQualityTools:
     # ===========================================
     
     def _build_full_table_name(self, table_name: str, catalog_name: Optional[str], db_name: Optional[str]) -> str:
-        """Build full table name"""
-        if catalog_name and db_name:
-            return f"{catalog_name}.{db_name}.{table_name}"
-        elif db_name:
-            return f"{db_name}.{table_name}"
-        else:
-            return table_name
+        """Build full table name with security validation"""
+        # SECURITY FIX: Use build_table_reference for safe identifier handling
+        return build_table_reference(table_name, db_name, catalog_name)
     
     async def _get_table_basic_info(self, connection, table_name: str) -> Optional[Dict]:
         """Get basic table information"""
         try:
+            # SECURITY FIX: table_name should already be validated by _build_full_table_name
+            # But we add auth_context for security validation
+            auth_context = get_auth_context()
+            
             # Try to get row count
             count_sql = f"SELECT COUNT(*) as row_count FROM {table_name}"
-            result = await connection.execute(count_sql)
+            result = await connection.execute(count_sql, auth_context=auth_context)
             if result.data:
                 return {"row_count": result.data[0]["row_count"]}
             return None
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed: {str(e)}")
+            return None
         except Exception as e:
             logger.warning(f"Failed to get table basic info: {str(e)}")
             return None
@@ -323,9 +332,13 @@ class DataQualityTools:
     async def _get_table_columns_info(self, connection, table_name: str, catalog_name: Optional[str], db_name: Optional[str]) -> List[Dict]:
         """Get table column information"""
         try:
-            # Build DESCRIBE query
-            describe_sql = f"DESCRIBE {self._build_full_table_name(table_name, catalog_name, db_name)}"
-            result = await connection.execute(describe_sql)
+            # SECURITY FIX: Build safe table reference and pass auth_context
+            auth_context = get_auth_context()
+            
+            # Build DESCRIBE query with safe table reference
+            safe_table_ref = self._build_full_table_name(table_name, catalog_name, db_name)
+            describe_sql = f"DESCRIBE {safe_table_ref}"
+            result = await connection.execute(describe_sql, auth_context=auth_context)
             
             columns_info = []
             if result.data:
@@ -339,6 +352,9 @@ class DataQualityTools:
                     })
             
             return columns_info
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed: {str(e)}")
+            return []
         except Exception as e:
             logger.warning(f"Failed to get table columns info: {str(e)}")
             return []
@@ -346,7 +362,32 @@ class DataQualityTools:
     async def _get_table_partitions(self, connection, table_name: str, db_name: Optional[str] = None) -> List[Dict]:
         """Get table partition information"""
         try:
-            # Query partition information
+            # SECURITY FIX: Validate identifiers and use parameterized query
+            auth_context = get_auth_context()
+            
+            # Validate table_name
+            try:
+                validate_identifier(table_name, "table name")
+                if db_name:
+                    validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            # Build parameterized query
+            params = []
+            where_conditions = []
+            
+            if db_name:
+                where_conditions.append("TABLE_SCHEMA = %s")
+                params.append(db_name)
+            else:
+                where_conditions.append("TABLE_SCHEMA = ''")
+            
+            where_conditions.append("TABLE_NAME = %s")
+            params.append(table_name)
+            where_conditions.append("PARTITION_NAME IS NOT NULL")
+            
             partition_sql = f"""
             SELECT 
                 PARTITION_NAME,
@@ -355,12 +396,10 @@ class DataQualityTools:
                 DATA_LENGTH,
                 INDEX_LENGTH
             FROM information_schema.PARTITIONS 
-            WHERE TABLE_SCHEMA = '{db_name or ""}' 
-            AND TABLE_NAME = '{table_name}'
-            AND PARTITION_NAME IS NOT NULL
+            WHERE {' AND '.join(where_conditions)}
             """
             
-            result = await connection.execute(partition_sql)
+            result = await connection.execute(partition_sql, params=tuple(params), auth_context=auth_context)
             partitions = []
             if result.data:
                 for row in result.data:
@@ -373,6 +412,9 @@ class DataQualityTools:
                     })
             
             return partitions
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed: {str(e)}")
+            return []
         except Exception as e:
             logger.warning(f"Failed to get table partitions: {str(e)}")
             return []
@@ -417,7 +459,8 @@ class DataQualityTools:
                 if db_name
                 else f"SHOW CREATE TABLE {table_name}"
             )
-            result = await connection.execute(query)
+            auth_context = get_auth_context()
+            result = await connection.execute(query, auth_context=auth_context)
             if result.data:
                 return result.data[0].get("Create Table")
             return None
@@ -428,8 +471,16 @@ class DataQualityTools:
     async def _get_table_size_info(self, connection, table_name: str) -> Dict[str, Any]:
         """Get table size information"""
         try:
-            # Query table size information
-            size_sql = f"""
+            # SECURITY FIX: Validate and use parameterized query
+            table_name_part = table_name.split('.')[-1]
+            try:
+                validate_identifier(table_name_part, "table name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid table name rejected: {e}")
+                return {"engine": "Unknown", "estimated_rows": 0, "data_length": 0, "index_length": 0, "total_size": 0}
+            
+            auth_context = get_auth_context()
+            size_sql = """
             SELECT 
                 table_name,
                 engine,
@@ -438,10 +489,10 @@ class DataQualityTools:
                 index_length,
                 (data_length + index_length) as total_size
             FROM information_schema.tables 
-            WHERE table_name = '{table_name.split('.')[-1]}'
+            WHERE table_name = %s
             """
             
-            result = await connection.execute(size_sql)
+            result = await connection.execute(size_sql, params=(table_name_part,), auth_context=auth_context)
             if result.data and result.data[0]:
                 row = result.data[0]
                 return {
@@ -582,7 +633,8 @@ class DataQualityTools:
             
             batch_sql = f"SELECT {', '.join(select_clauses)} FROM {table_expr}"
             
-            result = await connection.execute(batch_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(batch_sql, auth_context=auth_context)
             if not result.data:
                 return {"error": "No data returned from batch completeness query"}
             
@@ -664,7 +716,8 @@ class DataQualityTools:
             
             batch_sql = f"SELECT {', '.join(select_clauses)} FROM {table_expr}"
             
-            result = await connection.execute(batch_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(batch_sql, auth_context=auth_context)
             if not result.data:
                 return {}
             
@@ -705,7 +758,8 @@ class DataQualityTools:
                 LIMIT 10
                 """
                 
-                result = await connection.execute(freq_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(freq_sql, auth_context=auth_context)
                 frequencies = result.data if result.data else []
                 
                 categorical_results[col_name] = {
@@ -738,7 +792,8 @@ class DataQualityTools:
             
             batch_sql = f"SELECT {', '.join(select_clauses)} FROM {table_expr}"
             
-            result = await connection.execute(batch_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(batch_sql, auth_context=auth_context)
             if not result.data:
                 return {}
             
@@ -780,7 +835,8 @@ class DataQualityTools:
                 FROM {table_expr}
                 """
                 
-                result = await connection.execute(completeness_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(completeness_sql, auth_context=auth_context)
                 if result.data:
                     stats = result.data[0]
                     total_count = stats["total_count"]
@@ -906,7 +962,8 @@ class DataQualityTools:
                 WHERE {col_name} IS NOT NULL
                 """
                 
-                result = await connection.execute(stats_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(stats_sql, auth_context=auth_context)
                 if result.data and result.data[0]["non_null_count"] > 0:
                     stats = result.data[0]
                     numeric_analysis[col_name] = {
@@ -945,7 +1002,8 @@ class DataQualityTools:
                 WHERE {col_name} IS NOT NULL
                 """
                 
-                cardinality_result = await connection.execute(cardinality_sql)
+                auth_context = get_auth_context()
+                cardinality_result = await connection.execute(cardinality_sql, auth_context=auth_context)
                 
                 if cardinality_result.data:
                     stats = cardinality_result.data[0]
@@ -969,7 +1027,7 @@ class DataQualityTools:
                         LIMIT 10
                         """
                         
-                        top_values_result = await connection.execute(top_values_sql)
+                        top_values_result = await connection.execute(top_values_sql, auth_context=auth_context)
                         if top_values_result.data:
                             categorical_analysis[col_name]["top_values"] = [
                                 {"value": row[col_name], "count": row["count"]}
@@ -998,7 +1056,8 @@ class DataQualityTools:
                 WHERE {col_name} IS NOT NULL
                 """
                 
-                result = await connection.execute(stats_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(stats_sql, auth_context=auth_context)
                 if result.data and result.data[0]["non_null_count"] > 0:
                     stats = result.data[0]
                     temporal_analysis[col_name] = {

doris_mcp_server/utils/db.py

@@ -59,6 +59,7 @@ class QueryResult:
     metadata: dict[str, Any]
     execution_time: float
     row_count: int
+    sql: str
 
 
 class DorisConnection:
@@ -95,12 +96,14 @@ class DorisConnection:
                 await cursor.execute(sql, params)
 
                 # Check if it's a query statement (statement that returns result set)
+                # FIX for Issue #62 Bug 5: Added WITH support for Common Table Expressions (CTE)
                 sql_upper = sql.strip().upper()
                 if (sql_upper.startswith("SELECT") or
                     sql_upper.startswith("SHOW") or
                     sql_upper.startswith("DESCRIBE") or
                     sql_upper.startswith("DESC") or
-                    sql_upper.startswith("EXPLAIN")):
+                    sql_upper.startswith("EXPLAIN") or
+                    sql_upper.startswith("WITH")):  # FIX: Support CTE queries
                     data = await cursor.fetchall()
                     row_count = len(data)
                 else:
@@ -130,6 +133,7 @@ class DorisConnection:
                     metadata=metadata,
                     execution_time=execution_time,
                     row_count=row_count,
+                    sql=sql
                 )
 
         except Exception as e:
@@ -250,7 +254,23 @@ class DorisConnectionManager:
         self.logger = get_logger(__name__)
         self.security_manager = security_manager
         self.token_manager = token_manager  # Token manager for token-bound DB config
-        self.session_cache = DorisSessionCache(self)
+
+        # 🔧 FIX for multi-tenant concurrency: Per-token connection pool isolation
+        # Each token gets its own connection pool to prevent configuration conflicts
+        self.token_pools: Dict[str, Pool] = {}  # token_hash -> pool
+        self.token_configs: Dict[str, dict] = {}  # token_hash -> db_config
+        self._token_pool_locks: Dict[str, asyncio.Lock] = {}  # token_hash -> lock
+        self._token_pools_lock = asyncio.Lock()  # Lock for managing token_pools dict
+
+        # FIX for Issue #58 Problem 1: Disable session caching to prevent connection sharing
+        # Session caching causes multiple threads to share the same MySQL connection,
+        # leading to race conditions and deadlocks in multi-threaded environments
+        # By disabling caching, each request gets a fresh connection from the pool
+        self.session_cache = DorisSessionCache(
+            self,
+            cache_system_session=False,  # Disabled to prevent multi-thread issues
+            cache_user_session=False     # Disabled to prevent multi-thread issues
+        )
         
         # Store original database config for fallback
         self.original_db_config = {
@@ -263,6 +283,7 @@ class DorisConnectionManager:
         }
         
         # Current active database config (may be overridden by token-bound config)
+        # NOTE: This is kept for backward compatibility with non-token requests
         self.active_db_config = self.original_db_config.copy()
 
         # Connection pool state management
@@ -346,6 +367,281 @@ class DorisConnectionManager:
             self.logger.error(f"Error finding available token: {e}")
             return ""
     
+    def _get_token_hash(self, token: str) -> str:
+        """Get hash of token for use as dictionary key"""
+        import hashlib
+        return hashlib.sha256(token.encode()).hexdigest()[:16]
+    
+    def _get_current_token_db_config(self, token: str) -> dict | None:
+        """Get current database config for token from TokenManager
+        
+        This is used to check if config has changed for hot reload support.
+        """
+        if not self.token_manager:
+            return None
+        
+        token_db_config = self.token_manager.get_database_config_by_token(token)
+        if token_db_config:
+            return {
+                'host': token_db_config.host,
+                'port': token_db_config.port,
+                'user': token_db_config.user,
+                'password': token_db_config.password,
+                'database': token_db_config.database,
+                'charset': token_db_config.charset
+            }
+        return None
+    
+    def _config_changed(self, old_config: dict, new_config: dict) -> bool:
+        """Check if database configuration has changed"""
+        if old_config is None or new_config is None:
+            return old_config != new_config
+        
+        # Compare key fields
+        for key in ['host', 'port', 'user', 'password', 'database']:
+            if old_config.get(key) != new_config.get(key):
+                return True
+        return False
+    
+    async def get_pool_for_token(self, token: str) -> tuple[Pool, dict]:
+        """Get or create a dedicated connection pool for a specific token
+        
+        This method implements per-token connection pool isolation to prevent
+        concurrent requests from different tokens interfering with each other.
+        
+        🔧 FIX: Supports hot reload - if tokens.json config changes,
+        the old pool is closed and a new one is created automatically.
+        
+        Args:
+            token: Authentication token
+            
+        Returns:
+            (pool, db_config): The dedicated pool and its configuration
+            
+        Raises:
+            RuntimeError: If no valid database configuration is available
+        """
+        token_hash = self._get_token_hash(token)
+        
+        # Fast path: pool already exists
+        if token_hash in self.token_pools:
+            pool = self.token_pools[token_hash]
+            cached_config = self.token_configs.get(token_hash)
+            
+            # 🔧 FIX: Check if config has changed (hot reload support)
+            current_config = self._get_current_token_db_config(token)
+            if current_config and cached_config and self._config_changed(cached_config, current_config):
+                self.logger.info(f"Token config changed (hash: {token_hash[:8]}...), recreating pool...")
+                # Config changed, need to recreate pool
+                async with self._token_pools_lock:
+                    # Close old pool
+                    old_pool = self.token_pools.pop(token_hash, None)
+                    if old_pool and not old_pool.closed:
+                        try:
+                            old_pool.close()
+                            await asyncio.wait_for(old_pool.wait_closed(), timeout=2.0)
+                        except Exception as e:
+                            self.logger.warning(f"Error closing old pool during hot reload: {e}")
+                    self.token_configs.pop(token_hash, None)
+                # Continue to slow path to create new pool
+            elif pool and not pool.closed:
+                return pool, cached_config
+        
+        # Slow path: need to create pool (with lock to prevent race conditions)
+        async with self._token_pools_lock:
+            # Double-check after acquiring lock
+            if token_hash in self.token_pools:
+                pool = self.token_pools[token_hash]
+                if pool and not pool.closed:
+                    return pool, self.token_configs[token_hash]
+            
+            # Get database config for this token
+            db_config = None
+            config_source = "unknown"
+            
+            if self.token_manager:
+                token_db_config = self.token_manager.get_database_config_by_token(token)
+                if token_db_config:
+                    db_config = {
+                        'host': token_db_config.host,
+                        'port': token_db_config.port,
+                        'user': token_db_config.user,
+                        'password': token_db_config.password,
+                        'database': token_db_config.database,
+                        'charset': token_db_config.charset
+                    }
+                    config_source = "token-bound"
+            
+            # Fallback to global config if token has no specific config
+            if not db_config or self._is_config_empty(db_config.get('host')) or self._is_config_empty(db_config.get('user')):
+                if self._has_valid_global_config():
+                    db_config = self.original_db_config.copy()
+                    config_source = "global-env"
+                else:
+                    raise RuntimeError(
+                        f"No valid database configuration available for token. "
+                        f"Please configure database in tokens.json or .env file."
+                    )
+            
+            # Create dedicated pool for this token
+            self.logger.info(f"Creating dedicated connection pool for token (hash: {token_hash[:8]}...) "
+                           f"using {config_source} config: {db_config['user']}@{db_config['host']}:{db_config['port']}")
+            
+            pool = await self._create_pool_with_config(db_config)
+            
+            # Store pool and config
+            self.token_pools[token_hash] = pool
+            self.token_configs[token_hash] = db_config
+            
+            # Create lock for this token if not exists
+            if token_hash not in self._token_pool_locks:
+                self._token_pool_locks[token_hash] = asyncio.Lock()
+            
+            return pool, db_config
+    
+    async def _create_pool_with_config(self, db_config: dict) -> Pool:
+        """Create a connection pool with specified configuration
+        
+        Args:
+            db_config: Database configuration dictionary
+            
+        Returns:
+            Created connection pool
+        """
+        # Convert charset to aiomysql compatible format
+        charset_map = {"UTF8": "utf8", "UTF8MB4": "utf8mb4"}
+        charset = charset_map.get(db_config['charset'].upper(), db_config['charset'].lower())
+        
+        self.logger.debug(f"Creating pool for {db_config['user']}@{db_config['host']}:{db_config['port']}/{db_config['database']}")
+        
+        try:
+            pool = await asyncio.wait_for(
+                aiomysql.create_pool(
+                    host=db_config['host'],
+                    port=db_config['port'],
+                    user=db_config['user'],
+                    password=db_config['password'],
+                    db=db_config['database'],
+                    charset=charset,
+                    minsize=0,  # Don't pre-create connections
+                    maxsize=self.maxsize,
+                    connect_timeout=self.connect_timeout,
+                    autocommit=True,
+                    pool_recycle=self.pool_recycle
+                ),
+                timeout=self.connect_timeout + 5  # Give extra time for pool creation
+            )
+            self.logger.info(f"Successfully created pool for {db_config['user']}@{db_config['host']}:{db_config['port']}")
+            return pool
+        except asyncio.TimeoutError:
+            self.logger.error(f"Timeout creating pool for {db_config['user']}@{db_config['host']}:{db_config['port']}")
+            raise RuntimeError(f"Timeout creating connection pool for {db_config['user']}@{db_config['host']}:{db_config['port']}")
+        except Exception as e:
+            self.logger.error(f"Failed to create pool for {db_config['user']}@{db_config['host']}:{db_config['port']}: {type(e).__name__}: {e}")
+            raise
+    
+    async def get_connection_for_token(self, token: str, session_id: str) -> 'DorisConnection':
+        """Get a connection from the token's dedicated pool
+        
+        Args:
+            token: Authentication token
+            session_id: Session identifier for logging
+            
+        Returns:
+            DorisConnection wrapper
+        """
+        pool, db_config = await self.get_pool_for_token(token)
+        
+        try:
+            connection = await asyncio.wait_for(
+                pool.acquire(),
+                timeout=self.connect_timeout
+            )
+            
+            self.logger.debug(f"Session {session_id}: Acquired connection from token pool "
+                            f"(user: {db_config['user']}@{db_config['host']})")
+            
+            return DorisConnection(connection, session_id, self.security_manager)
+            
+        except Exception as e:
+            self.logger.error(f"Session {session_id}: Failed to acquire connection from token pool: {e}")
+            raise
+    
+    async def release_connection_for_token(self, token: str, connection: 'DorisConnection'):
+        """Release a connection back to the token's dedicated pool
+        
+        Args:
+            token: Authentication token
+            connection: DorisConnection wrapper to release
+        """
+        token_hash = self._get_token_hash(token)
+        
+        if token_hash in self.token_pools:
+            pool = self.token_pools[token_hash]
+            if pool and not pool.closed:
+                try:
+                    pool.release(connection.connection)
+                except Exception as e:
+                    self.logger.warning(f"Failed to release connection to token pool: {e}")
+    
+    async def cleanup_token_pools(self, max_idle_time: int = 3600):
+        """Clean up idle token connection pools
+        
+        Args:
+            max_idle_time: Maximum idle time in seconds before closing a pool
+        """
+        async with self._token_pools_lock:
+            pools_to_remove = []
+            
+            for token_hash, pool in self.token_pools.items():
+                if pool and not pool.closed:
+                    # Check if pool is idle (no active connections)
+                    if pool.size == 0 and pool.freesize == 0:
+                        pools_to_remove.append(token_hash)
+                elif pool and pool.closed:
+                    pools_to_remove.append(token_hash)
+            
+            for token_hash in pools_to_remove:
+                try:
+                    pool = self.token_pools.pop(token_hash, None)
+                    if pool and not pool.closed:
+                        pool.close()
+                        await pool.wait_closed()
+                    self.token_configs.pop(token_hash, None)
+                    self._token_pool_locks.pop(token_hash, None)
+                    self.logger.info(f"Cleaned up idle token pool (hash: {token_hash[:8]}...)")
+                except Exception as e:
+                    self.logger.warning(f"Error cleaning up token pool: {e}")
+    
+    async def close_all_token_pools(self):
+        """Close all token connection pools (for shutdown)"""
+        # Use timeout to prevent blocking on lock acquisition during shutdown
+        try:
+            async with asyncio.timeout(5):  # 5 second timeout for lock
+                async with self._token_pools_lock:
+                    for token_hash, pool in list(self.token_pools.items()):
+                        try:
+                            if pool and not pool.closed:
+                                pool.close()
+                                # Use timeout for wait_closed to prevent hanging
+                                try:
+                                    await asyncio.wait_for(pool.wait_closed(), timeout=2.0)
+                                except asyncio.TimeoutError:
+                                    self.logger.warning(f"Timeout waiting for token pool to close (hash: {token_hash[:8]}...)")
+                                self.logger.info(f"Closed token pool (hash: {token_hash[:8]}...)")
+                        except Exception as e:
+                            self.logger.warning(f"Error closing token pool: {e}")
+                    
+                    self.token_pools.clear()
+                    self.token_configs.clear()
+                    self._token_pool_locks.clear()
+        except asyncio.TimeoutError:
+            self.logger.warning("Timeout acquiring lock for token pool cleanup, forcing clear")
+            # Force clear without lock
+            self.token_pools.clear()
+            self.token_configs.clear()
+            self._token_pool_locks.clear()
+
     async def configure_for_token(self, token: str) -> tuple[bool, str]:
         """Configure connection manager for token with new priority logic
         
@@ -626,6 +922,213 @@ class DorisConnectionManager:
             self.logger.error(f"Failed to initialize connection pool: {e}")
             raise
 
+    async def initialize_for_stdio_mode(self, timeout: float = 30.0) -> None:
+        """
+        Initialize connection pool for stdio mode with strict validation
+        
+        stdio mode requires a working database connection because:
+        - No HTTP authentication mechanism to support token-bound configs
+        - All database operations depend on the global connection pool
+        
+        Args:
+            timeout: Maximum time to wait for connection establishment
+            
+        Raises:
+            RuntimeError: If configuration is invalid or connection fails
+        """
+        try:
+            # Validate that we have valid global configuration
+            if not self._has_valid_global_config():
+                error_msg = (
+                    "stdio mode requires valid global database configuration. "
+                    "Please set DORIS_HOST and DORIS_USER in environment variables or .env file. "
+                    f"Current config: host='{self.host}', user='{self.user}'"
+                )
+                self.logger.error(error_msg)
+                raise RuntimeError(error_msg)
+            
+            self.logger.info(f"stdio mode database config validated: {self.host}:{self.port}")
+            
+            # Validate configuration format
+            is_valid, error_message = self.validate_database_configuration()
+            if not is_valid:
+                error_msg = f"Database configuration validation failed: {error_message}"
+                self.logger.error(error_msg)
+                raise RuntimeError(error_msg)
+            
+            # Test connectivity with timeout
+            self.logger.info("Testing database connectivity for stdio mode...")
+            if not await self._test_connectivity_with_timeout(timeout):
+                error_msg = (
+                    f"Failed to connect to Doris database within {timeout} seconds. "
+                    f"Please check if Doris is running at {self.host}:{self.port} "
+                    f"and verify network connectivity."
+                )
+                self.logger.error(error_msg)
+                raise RuntimeError(error_msg)
+            
+            # Initialize the connection pool
+            await self._create_connection_pool()
+            
+            # Verify that we have a working connection pool
+            if not self.pool:
+                error_msg = "Database connection pool was not created successfully."
+                self.logger.error(error_msg)
+                raise RuntimeError(error_msg)
+            
+            # Start background monitoring tasks
+            self.pool_health_check_task = asyncio.create_task(self._pool_health_monitor())
+            self.pool_cleanup_task = asyncio.create_task(self._pool_cleanup_monitor())
+            
+            # Perform initial pool warmup
+            await self._warmup_pool()
+            
+            self.logger.info("Database connection established successfully for stdio mode")
+            
+        except Exception as e:
+            self.logger.error(f"stdio mode database initialization failed: {e}")
+            raise
+    
+    async def initialize_for_http_mode(self) -> bool:
+        """
+        Initialize connection pool for HTTP mode with graceful degradation
+        
+        HTTP mode can work without global database configuration because:
+        - Supports token-bound database configurations
+        - Can handle authentication and use per-request database configs
+        - Has fallback mechanisms for database operations
+        
+        Returns:
+            bool: True if global database pool was created, False if gracefully degraded
+        """
+        try:
+            # First validate configuration format if we have one
+            if self._has_valid_global_config():
+                is_valid, error_message = self.validate_database_configuration()
+                if not is_valid:
+                    self.logger.warning(f"Global database configuration invalid: {error_message}")
+                    self.logger.info("HTTP mode will rely on token-bound database configurations")
+                    return False
+                
+                # Try to establish global connection pool
+                self.logger.info(f"Attempting to create global connection pool: {self.host}:{self.port}")
+                
+                try:
+                    # Test connectivity with shorter timeout for HTTP mode
+                    if await self._test_connectivity_with_timeout(10.0):
+                        await self._create_connection_pool()
+                        
+                        if self.pool:
+                            # Start background monitoring tasks
+                            self.pool_health_check_task = asyncio.create_task(self._pool_health_monitor())
+                            self.pool_cleanup_task = asyncio.create_task(self._pool_cleanup_monitor())
+                            
+                            # Perform initial pool warmup
+                            await self._warmup_pool()
+                            
+                            self.logger.info("Global database connection pool created successfully for HTTP mode")
+                            return True
+                    else:
+                        self.logger.warning("Global database connection test failed, will use token-bound configs")
+                        return False
+                        
+                except Exception as pool_error:
+                    self.logger.warning(f"Failed to create global connection pool: {pool_error}")
+                    self.logger.info("HTTP mode will rely on token-bound database configurations")
+                    return False
+            else:
+                self.logger.info("No valid global database config found, HTTP mode will use token-bound configurations")
+                return False
+                
+        except Exception as e:
+            self.logger.warning(f"HTTP mode database initialization encountered error: {e}")
+            self.logger.info("HTTP mode will rely on token-bound database configurations")
+            return False
+    
+    async def _test_connectivity_with_timeout(self, timeout: float) -> bool:
+        """
+        Test database connectivity with timeout
+        
+        Args:
+            timeout: Maximum time to wait for connection test
+            
+        Returns:
+            bool: True if connection successful, False otherwise
+        """
+        try:
+            await asyncio.wait_for(self._test_basic_connectivity(), timeout=timeout)
+            return True
+        except asyncio.TimeoutError:
+            self.logger.error(f"Database connectivity test timed out after {timeout} seconds")
+            return False
+        except Exception as e:
+            self.logger.error(f"Database connectivity test failed: {e}")
+            return False
+    
+    async def _test_basic_connectivity(self) -> None:
+        """
+        Test basic database connectivity without connection pool
+        
+        Raises:
+            Exception: If connection fails
+        """
+        import aiomysql
+        
+        conn = None
+        try:
+            conn = await aiomysql.connect(
+                host=self.host,
+                port=self.port,
+                user=self.user,
+                password=self.password,
+                db=self.database,
+                charset=self.charset,
+                connect_timeout=self.connect_timeout,
+                autocommit=True
+            )
+            
+            async with conn.cursor() as cursor:
+                await cursor.execute("SELECT 1")
+                result = await cursor.fetchone()
+                if not result or result[0] != 1:
+                    raise RuntimeError("Database connectivity test query failed")
+                    
+        except Exception as e:
+            raise RuntimeError(f"Database connectivity test failed: {e}")
+        finally:
+            if conn:
+                conn.close()
+    
+    async def _create_connection_pool(self) -> None:
+        """
+        Create the connection pool
+        
+        Raises:
+            Exception: If pool creation fails
+        """
+        self.pool = await aiomysql.create_pool(
+            host=self.host,
+            port=self.port,
+            user=self.user,
+            password=self.password,
+            db=self.database,
+            charset=self.charset,
+            minsize=self.minsize,
+            maxsize=self.maxsize,
+            pool_recycle=self.pool_recycle,
+            connect_timeout=self.connect_timeout,
+            autocommit=True
+        )
+        
+        # Test pool health
+        if not await self._test_pool_health():
+            # Clean up the pool if health test fails
+            if self.pool:
+                self.pool.close()
+                await self.pool.wait_closed()
+                self.pool = None
+            raise RuntimeError("Connection pool health check failed")
+
     async def _test_pool_health(self) -> bool:
         """Test connection pool health"""
         try:
@@ -872,7 +1375,26 @@ class DorisConnectionManager:
         
         Uses only semaphore to prevent too many concurrent acquisitions.
         If the connection is successfully obtained, it will be added to the connection pool cache.
+        
+        🔧 FIX for token isolation: Now automatically checks for auth_context from ContextVar
+        and uses token-specific connection pool if available.
         """
+        # 🔧 FIX: Check for auth_context from global ContextVar
+        # This ensures all tools using get_connection respect token-bound database configuration
+        auth_context = None
+        try:
+            from .security import mcp_auth_context_var
+            auth_context = mcp_auth_context_var.get()
+        except Exception as e:
+            self.logger.debug(f"get_connection: Could not get auth_context: {e}")
+        
+        if auth_context and hasattr(auth_context, 'token') and auth_context.token:
+            # Use token-specific connection pool
+            # SECURITY: Do NOT catch exceptions here - if token pool fails, don't fallback to global pool
+            # This prevents privilege escalation
+            self.logger.debug(f"get_connection: Using token-specific pool for session {session_id}")
+            return await self.get_connection_for_token(auth_context.token, session_id)
+        
         cached_conn = self.session_cache.get(session_id)
         if cached_conn:
             return cached_conn
@@ -1019,10 +1541,16 @@ class DorisConnectionManager:
                 except asyncio.CancelledError:
                     pass
 
-            # Close connection pool
+            # 🔧 FIX: Close all per-token connection pools
+            await self.close_all_token_pools()
+
+            # Close global connection pool with timeout
             if self.pool:
                 self.pool.close()
-                await self.pool.wait_closed()
+                try:
+                    await asyncio.wait_for(self.pool.wait_closed(), timeout=5.0)
+                except asyncio.TimeoutError:
+                    self.logger.warning("Timeout waiting for global pool to close")
 
             self.logger.info("Connection manager closed successfully")
 
@@ -1051,10 +1579,42 @@ class DorisConnectionManager:
     async def execute_query(
         self, session_id: str, sql: str, params: tuple | None = None, auth_context=None
     ) -> QueryResult:
-        """Execute query - Simplified Strategy with automatic connection management"""
+        """Execute query - Enhanced Strategy with per-token connection pool isolation
+
+        FIX for multi-tenant concurrency: Each token now uses its own dedicated connection pool
+        to prevent configuration conflicts between concurrent requests from different tokens.
+        """
         connection = None
+        token = None
+        
         try:
-            # Always get fresh connection from pool
+            # Check if we have a token for per-token pool isolation
+            if auth_context and hasattr(auth_context, 'token') and auth_context.token:
+                token = auth_context.token
+                
+                try:
+                    # 🔧 FIX: Use dedicated connection pool for this token
+                    # This prevents concurrent requests from different tokens interfering
+                    connection = await self.get_connection_for_token(token, session_id)
+                    
+                    # Get the config for logging
+                    token_hash = self._get_token_hash(token)
+                    if token_hash in self.token_configs:
+                        db_config = self.token_configs[token_hash]
+                        self.logger.info(f"Session {session_id}: Using dedicated pool for {db_config['user']}@{db_config['host']}")
+                    
+                except Exception as token_pool_error:
+                    # SECURITY: If token should have pool but creation fails, don't fallback
+                    # This prevents privilege escalation (using high-privilege default user)
+                    self.logger.error(f"Session {session_id}: Token pool error: {token_pool_error}")
+                    raise RuntimeError(
+                        f"Failed to get connection for authenticated token. "
+                        f"This is a security measure to prevent using default high-privilege credentials. "
+                        f"Error: {token_pool_error}"
+                    )
+            else:
+                # No token - use global pool (backward compatibility)
+                self.logger.debug(f"Session {session_id}: No token, using global connection pool")
                 connection = await self.get_connection(session_id)
 
             # Execute query
@@ -1066,8 +1626,11 @@ class DorisConnectionManager:
             self.logger.error(f"Query execution failed for session {session_id}: {e}")
             raise
         finally:
-            # Always release connection back to pool
+            # Always release connection back to the appropriate pool
             if connection:
+                if token:
+                    await self.release_connection_for_token(token, connection)
+                else:
                     await self.release_connection(session_id, connection)
 
     @asynccontextmanager

doris_mcp_server/utils/dependency_analysis_tools.py

@@ -27,6 +27,13 @@ from collections import defaultdict, deque
 
 from .db import DorisConnectionManager
 from .logger import get_logger
+from .sql_security_utils import (
+    SQLSecurityError,
+    validate_identifier,
+    quote_identifier,
+    build_table_reference,
+    get_auth_context
+)
 
 logger = get_logger(__name__)
 
@@ -122,10 +129,19 @@ class DependencyAnalysisTools:
     async def _get_tables_metadata(self, connection, catalog_name: Optional[str], db_name: Optional[str], include_views: bool) -> List[Dict]:
         """Get metadata for all tables and views"""
         try:
-            # Build conditions for query
+            # Build conditions for query with parameterized values
             where_conditions = []
+            params = []
+            
             if db_name:
-                where_conditions.append(f"table_schema = '{db_name}'")
+                # SECURITY FIX: Validate identifier and use parameterized query
+                try:
+                    validate_identifier(db_name, "database name")
+                except SQLSecurityError as e:
+                    logger.warning(f"Invalid database name rejected: {e}")
+                    return []
+                where_conditions.append("table_schema = %s")
+                params.append(db_name)
             else:
                 where_conditions.append("table_schema = DATABASE()")
             
@@ -148,9 +164,18 @@ class DependencyAnalysisTools:
             ORDER BY table_schema, table_name
             """
             
-            result = await connection.execute(metadata_sql)
+            # SECURITY FIX: Get auth_context and pass to execute for security validation
+            auth_context = get_auth_context()
+            result = await connection.execute(
+                metadata_sql, 
+                params=tuple(params) if params else None,
+                auth_context=auth_context
+            )
             return result.data if result.data else []
             
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed in _get_tables_metadata: {str(e)}")
+            return []
         except Exception as e:
             logger.warning(f"Failed to get tables metadata: {str(e)}")
             return []
@@ -186,17 +211,31 @@ class DependencyAnalysisTools:
     
     async def _analyze_view_dependencies(self, connection, dependency_graph: Dict, tables_metadata: List[Dict]) -> None:
         """Analyze view definitions to extract table dependencies"""
+        # Get auth_context once for all operations in this method
+        auth_context = get_auth_context()
+        
         try:
             for table in tables_metadata:
                 if table["table_type"] == "VIEW":
                     table_name = table["table_name"]
                     schema_name = table.get("schema_name", "")
                     
-                    # Get view definition
-                    view_def_sql = f"SHOW CREATE VIEW {schema_name}.{table_name}" if schema_name else f"SHOW CREATE VIEW {table_name}"
+                    # SECURITY FIX: Validate identifiers before using in SQL
+                    try:
+                        validate_identifier(table_name, "table name")
+                        if schema_name:
+                            validate_identifier(schema_name, "schema name")
+                    except SQLSecurityError as e:
+                        logger.warning(f"Invalid identifier rejected in view analysis: {e}")
+                        continue
+                    
+                    # Build safe view reference using quoted identifiers
+                    view_ref = build_table_reference(table_name, schema_name) if schema_name else quote_identifier(table_name, "table name")
+                    view_def_sql = f"SHOW CREATE VIEW {view_ref}"
                     
                     try:
-                        result = await connection.execute(view_def_sql)
+                        # SECURITY FIX: Pass auth_context to execute
+                        result = await connection.execute(view_def_sql, auth_context=auth_context)
                         if result.data and len(result.data) > 0:
                             # Extract view definition from result
                             view_definition = ""
@@ -235,6 +274,9 @@ class DependencyAnalysisTools:
     
     async def _analyze_runtime_dependencies(self, connection, dependency_graph: Dict, analysis_depth: int) -> None:
         """Analyze audit logs to discover runtime table dependencies"""
+        # Get auth_context for security validation
+        auth_context = get_auth_context()
+        
         try:
             # Get recent SQL statements from audit logs
             audit_sql = """
@@ -252,7 +294,8 @@ class DependencyAnalysisTools:
             LIMIT 1000
             """
             
-            result = await connection.execute(audit_sql)
+            # SECURITY FIX: Pass auth_context to execute
+            result = await connection.execute(audit_sql, auth_context=auth_context)
             
             if result.data:
                 for row in result.data:
@@ -274,6 +317,9 @@ class DependencyAnalysisTools:
     
     async def _analyze_foreign_key_dependencies(self, connection, dependency_graph: Dict, tables_metadata: List[Dict]) -> None:
         """Analyze foreign key constraints for explicit dependencies"""
+        # Get auth_context for security validation
+        auth_context = get_auth_context()
+        
         try:
             # Get foreign key information
             fk_sql = """
@@ -288,7 +334,8 @@ class DependencyAnalysisTools:
             WHERE REFERENCED_TABLE_NAME IS NOT NULL
             """
             
-            result = await connection.execute(fk_sql)
+            # SECURITY FIX: Pass auth_context to execute
+            result = await connection.execute(fk_sql, auth_context=auth_context)
             
             if result.data:
                 for row in result.data:

doris_mcp_server/utils/monitoring_tools.py

@@ -28,6 +28,7 @@ from datetime import datetime
 
 from .db import DorisConnectionManager
 from .logger import get_logger
+from .sql_security_utils import get_auth_context
 
 logger = get_logger(__name__)
 
@@ -713,7 +714,8 @@ class DorisMonitoringTools:
             # Fallback to SHOW BACKENDS if no BE hosts configured
             logger.info("No BE hosts configured, using SHOW BACKENDS to discover BE nodes")
             connection = await self.connection_manager.get_connection("query")
-            result = await connection.execute("SHOW BACKENDS")
+            auth_context = get_auth_context()
+            result = await connection.execute("SHOW BACKENDS", auth_context=auth_context)
             
             be_nodes = []
             for row in result.data:

doris_mcp_server/utils/performance_analytics_tools.py

@@ -27,6 +27,13 @@ from collections import defaultdict, Counter
 
 from .db import DorisConnectionManager
 from .logger import get_logger
+from .sql_security_utils import (
+    SQLSecurityError,
+    validate_identifier,
+    quote_identifier,
+    build_table_reference,
+    get_auth_context
+)
 
 logger = get_logger(__name__)
 
@@ -229,7 +236,8 @@ class PerformanceAnalyticsTools:
             ORDER BY query_date
             """
             
-            result = await connection.execute(query_volume_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(query_volume_sql, auth_context=auth_context)
             daily_data = result.data if result.data else []
             
             if not daily_data:
@@ -304,7 +312,8 @@ class PerformanceAnalyticsTools:
             ORDER BY activity_date
             """
             
-            result = await connection.execute(user_activity_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(user_activity_sql, auth_context=auth_context)
             daily_data = result.data if result.data else []
             
             if not daily_data:
@@ -383,7 +392,8 @@ class PerformanceAnalyticsTools:
             LIMIT 5000
             """
             
-            result = await connection.execute(slow_query_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(slow_query_sql, auth_context=auth_context)
             return result.data if result.data else []
             
         except Exception as e:
@@ -705,7 +715,8 @@ class PerformanceAnalyticsTools:
             ORDER BY size_mb DESC
             """
             
-            db_result = await connection.execute(db_sizes_sql)
+            auth_context = get_auth_context()
+            db_result = await connection.execute(db_sizes_sql, auth_context=auth_context)
             
             if not db_result.data:
                 logger.warning("No database size information available")
@@ -805,7 +816,16 @@ class PerformanceAnalyticsTools:
     async def _get_database_table_details_from_schema(self, connection, db_name: str) -> List[Dict]:
         """Get table details for a specific database using information_schema"""
         try:
-            table_details_sql = f"""
+            # SECURITY FIX: Validate db_name and use parameterized query
+            auth_context = get_auth_context()
+            
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid database name rejected: {e}")
+                return []
+            
+            table_details_sql = """
             SELECT 
                 TABLE_SCHEMA as schema_name,
                 TABLE_NAME as table_name,
@@ -814,13 +834,13 @@ class PerformanceAnalyticsTools:
                 CREATE_TIME as create_time,
                 UPDATE_TIME as update_time
             FROM information_schema.tables 
-            WHERE TABLE_SCHEMA = '{db_name}'
+            WHERE TABLE_SCHEMA = %s
                 AND TABLE_TYPE = 'BASE TABLE'
                 AND (COALESCE(DATA_LENGTH, 0) + COALESCE(INDEX_LENGTH, 0)) > 0
             ORDER BY size_mb DESC
             """
             
-            result = await connection.execute(table_details_sql)
+            result = await connection.execute(table_details_sql, params=(db_name,), auth_context=auth_context)
             
             if not result.data:
                 logger.warning(f"No table details found for database {db_name}")
@@ -867,6 +887,13 @@ class PerformanceAnalyticsTools:
     async def _get_database_table_details(self, connection, db_name: str) -> List[Dict]:
         """Get table details for a specific database using session-consistent queries"""
         try:
+            # SECURITY FIX: Validate db_name before using in SQL
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid database name rejected: {e}")
+                return []
+            
             # Method 1: Try to use session-consistent approach with raw connection
             # This requires accessing the underlying connection to maintain session state
             
@@ -877,8 +904,9 @@ class PerformanceAnalyticsTools:
                 # Use raw connection to maintain session state
                 cursor = await raw_conn.cursor()
                 try:
-                    # Execute USE and SHOW DATA in the same session
-                    await cursor.execute(f"USE {db_name}")
+                    # SECURITY FIX: Use quoted identifier for USE statement
+                    quoted_db = quote_identifier(db_name, "database name")
+                    await cursor.execute(f"USE {quoted_db}")
                     await cursor.execute("SHOW DATA")
                     
                     result = await cursor.fetchall()
@@ -922,9 +950,19 @@ class PerformanceAnalyticsTools:
     async def _get_database_table_details_fallback(self, connection, db_name: str) -> List[Dict]:
         """Fallback method to get table details using individual queries"""
         try:
-            # Get all tables in the database
-            tables_sql = f"SHOW TABLES FROM {db_name}"
-            tables_result = await connection.execute(tables_sql)
+            # SECURITY FIX: Validate db_name and get auth_context
+            auth_context = get_auth_context()
+            
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid database name rejected: {e}")
+                return []
+            
+            # Get all tables in the database using quoted identifier
+            quoted_db = quote_identifier(db_name, "database name")
+            tables_sql = f"SHOW TABLES FROM {quoted_db}"
+            tables_result = await connection.execute(tables_sql, auth_context=auth_context)
             
             if not tables_result.data:
                 return []
@@ -934,9 +972,11 @@ class PerformanceAnalyticsTools:
                 table_name = table_row.get(f"Tables_in_{db_name}", "") or table_row.get("table_name", "")
                 if table_name:
                     try:
-                        # Use SHOW DATA FROM db.table for each table
-                        data_sql = f"SHOW DATA FROM {db_name}.{table_name}"
-                        data_result = await connection.execute(data_sql)
+                        # SECURITY FIX: Validate table_name and use safe reference
+                        validate_identifier(table_name, "table name")
+                        safe_table_ref = build_table_reference(table_name, db_name)
+                        data_sql = f"SHOW DATA FROM {safe_table_ref}"
+                        data_result = await connection.execute(data_sql, auth_context=auth_context)
                         
                         if data_result.data:
                             for row in data_result.data:
@@ -1036,6 +1076,7 @@ class PerformanceAnalyticsTools:
     async def _get_all_tables_info(self, connection) -> List[Dict]:
         """Get basic information for all tables (fallback method)"""
         try:
+            auth_context = get_auth_context()
             tables_sql = """
             SELECT 
                 table_schema,
@@ -1053,7 +1094,7 @@ class PerformanceAnalyticsTools:
             ORDER BY (data_length + index_length) DESC
             """
             
-            result = await connection.execute(tables_sql)
+            result = await connection.execute(tables_sql, auth_context=auth_context)
             return result.data if result.data else []
             
         except Exception as e:
@@ -1120,23 +1161,37 @@ class PerformanceAnalyticsTools:
     async def _get_current_table_size(self, connection, full_table_name: str) -> Optional[Dict]:
         """Get current table size"""
         try:
-            # Try to query table size directly
-            size_sql = f"""
+            # SECURITY FIX: Get auth_context and use parameterized query
+            auth_context = get_auth_context()
+            
+            # Extract table name for parameterized query
+            table_name_only = full_table_name.split('.')[-1] if '.' in full_table_name else full_table_name
+            
+            # Validate identifiers
+            try:
+                validate_identifier(table_name_only, "table name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid table name rejected: {e}")
+                return None
+            
+            # Use parameterized query for safety
+            size_sql = """
             SELECT 
                 COALESCE(ROUND((COALESCE(data_length, 0) + COALESCE(index_length, 0)) / 1024 / 1024, 2), 0) as size_mb,
                 COALESCE(table_rows, 0) as `rows`
             FROM information_schema.tables
-            WHERE CONCAT(table_schema, '.', table_name) = '{full_table_name}'
-                OR table_name = '{full_table_name.split('.')[-1]}'
+            WHERE CONCAT(table_schema, '.', table_name) = %s
+                OR table_name = %s
             """
             
-            result = await connection.execute(size_sql)
+            result = await connection.execute(size_sql, params=(full_table_name, table_name_only), auth_context=auth_context)
             if result.data and result.data[0]:
                 return result.data[0]
             
             # If information_schema has no data, try COUNT query
+            # full_table_name should already be validated by caller using build_table_reference
             count_sql = f"SELECT COUNT(*) as rows FROM {full_table_name}"
-            count_result = await connection.execute(count_sql)
+            count_result = await connection.execute(count_sql, auth_context=auth_context)
             if count_result.data:
                 return {
                     "size_mb": 0,  # Cannot get exact size
@@ -1145,6 +1200,9 @@ class PerformanceAnalyticsTools:
             
             return None
             
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed for {full_table_name}: {str(e)}")
+            return None
         except Exception as e:
             logger.warning(f"Failed to get current size for {full_table_name}: {str(e)}")
             return None
@@ -1154,8 +1212,19 @@ class PerformanceAnalyticsTools:
     ) -> List[Dict]:
         """Get historical growth data based on partitions"""
         try:
-            # Query partition information
-            partition_sql = f"""
+            # SECURITY FIX: Validate identifiers and use parameterized query
+            auth_context = get_auth_context()
+            
+            try:
+                validate_identifier(table_name, "table name")
+                if schema_name:
+                    validate_identifier(schema_name, "schema name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            # Use parameterized query for safety
+            partition_sql = """
             SELECT 
                 partition_name,
                 partition_description,
@@ -1163,15 +1232,19 @@ class PerformanceAnalyticsTools:
                 data_length,
                 create_time
             FROM information_schema.partitions
-            WHERE table_schema = '{schema_name or ""}'
-                AND table_name = '{table_name}'
+            WHERE table_schema = %s
+                AND table_name = %s
                 AND partition_name IS NOT NULL
                 AND create_time IS NOT NULL
-                AND create_time >= DATE_SUB(NOW(), INTERVAL {days} DAY)
+                AND create_time >= DATE_SUB(NOW(), INTERVAL %s DAY)
             ORDER BY create_time DESC
             """
             
-            result = await connection.execute(partition_sql)
+            result = await connection.execute(
+                partition_sql, 
+                params=(schema_name or "", table_name, days),
+                auth_context=auth_context
+            )
             if not result.data:
                 return []
             
@@ -1210,6 +1283,9 @@ class PerformanceAnalyticsTools:
     ) -> List[Dict]:
         """Get historical growth data based on timestamp fields"""
         try:
+            # SECURITY FIX: Get auth_context
+            auth_context = get_auth_context()
+            
             # Find possible timestamp fields
             timestamp_columns = await self._find_timestamp_columns(connection, table_name, schema_name)
             if not timestamp_columns:
@@ -1218,20 +1294,29 @@ class PerformanceAnalyticsTools:
             # Use best timestamp field for analysis
             time_column = timestamp_columns[0]
             
-            # Aggregate data by date
+            # SECURITY FIX: Validate time_column before using in SQL
+            try:
+                validate_identifier(time_column, "column name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid column name rejected: {e}")
+                return []
+            
+            quoted_time_column = quote_identifier(time_column, "column name")
+            
+            # Aggregate data by date (full_table_name should be validated by caller)
             growth_sql = f"""
             SELECT 
-                DATE({time_column}) as date,
+                DATE({quoted_time_column}) as date,
                 COUNT(*) as daily_records,
                 COUNT(*) / SUM(COUNT(*)) OVER() * 100 as percentage
             FROM {full_table_name}
-            WHERE {time_column} >= DATE_SUB(NOW(), INTERVAL {days} DAY)
-                AND {time_column} IS NOT NULL
-            GROUP BY DATE({time_column})
+            WHERE {quoted_time_column} >= DATE_SUB(NOW(), INTERVAL %s DAY)
+                AND {quoted_time_column} IS NOT NULL
+            GROUP BY DATE({quoted_time_column})
             ORDER BY date DESC
             """
             
-            result = await connection.execute(growth_sql)
+            result = await connection.execute(growth_sql, params=(days,), auth_context=auth_context)
             if not result.data:
                 return []
             
@@ -1257,11 +1342,22 @@ class PerformanceAnalyticsTools:
     async def _find_timestamp_columns(self, connection, table_name: str, schema_name: str) -> List[str]:
         """Find timestamp fields in table"""
         try:
-            timestamp_sql = f"""
+            # SECURITY FIX: Validate identifiers and use parameterized query
+            auth_context = get_auth_context()
+            
+            try:
+                validate_identifier(table_name, "table name")
+                if schema_name:
+                    validate_identifier(schema_name, "schema name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            timestamp_sql = """
             SELECT column_name, data_type
             FROM information_schema.columns
-            WHERE table_schema = '{schema_name or ""}'
-                AND table_name = '{table_name}'
+            WHERE table_schema = %s
+                AND table_name = %s
                 AND (
                     data_type IN ('datetime', 'timestamp', 'date')
                     OR column_name REGEXP '(create|insert|update|modify).*time'
@@ -1278,9 +1374,16 @@ class PerformanceAnalyticsTools:
                 END
             """
             
-            result = await connection.execute(timestamp_sql)
+            result = await connection.execute(
+                timestamp_sql, 
+                params=(schema_name or "", table_name),
+                auth_context=auth_context
+            )
             return [row["column_name"] for row in result.data] if result.data else []
             
+        except SQLSecurityError as e:
+            logger.warning(f"Security validation failed: {str(e)}")
+            return []
         except Exception as e:
             logger.warning(f"Failed to find timestamp columns: {str(e)}")
             return []
@@ -1290,8 +1393,22 @@ class PerformanceAnalyticsTools:
     ) -> List[Dict]:
         """Estimate growth data based on audit logs"""
         try:
+            # SECURITY FIX: Validate table_name and use parameterized query
+            auth_context = get_auth_context()
+            
+            try:
+                validate_identifier(table_name.split(".")[-1], "table name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid table name rejected: {e}")
+                return []
+            
+            # Extract just the table name for LIKE pattern
+            table_name_only = table_name.split(".")[-1]
+            like_pattern_full = f"%{table_name}%"
+            like_pattern_short = f"%{table_name_only}%"
+            
             # Analyze operation history for this table
-            audit_sql = f"""
+            audit_sql = """
             SELECT 
                 DATE(`time`) as operation_date,
                 COUNT(*) as operation_count,
@@ -1299,17 +1416,21 @@ class PerformanceAnalyticsTools:
                 SUM(CASE WHEN stmt LIKE 'UPDATE%' THEN 1 ELSE 0 END) as update_count,
                 SUM(CASE WHEN stmt LIKE 'DELETE%' THEN 1 ELSE 0 END) as delete_count
             FROM internal.__internal_schema.audit_log
-            WHERE `time` >= DATE_SUB(NOW(), INTERVAL {days} DAY)
+            WHERE `time` >= DATE_SUB(NOW(), INTERVAL %s DAY)
                 AND stmt IS NOT NULL
                 AND (
-                    stmt LIKE '%{table_name}%'
-                    OR stmt LIKE '%{table_name.split(".")[-1]}%'
+                    stmt LIKE %s
+                    OR stmt LIKE %s
                 )
             GROUP BY DATE(`time`)
             ORDER BY operation_date DESC
             """
             
-            result = await connection.execute(audit_sql)
+            result = await connection.execute(
+                audit_sql, 
+                params=(days, like_pattern_full, like_pattern_short),
+                auth_context=auth_context
+            )
             if not result.data:
                 return []
             

doris_mcp_server/utils/query_executor.py

@@ -33,8 +33,11 @@ from datetime import datetime, timedelta, date
 from typing import Any, Dict
 from decimal import Decimal
 
+import sqlparse
+
 from .db import DorisConnectionManager, QueryResult
 from .logger import get_logger
+from .sql_security_utils import get_auth_context
 
 
 @dataclass
@@ -467,6 +470,51 @@ class DorisQueryExecutor:
                 f"Slow query detected: {execution_time:.2f}s (threshold: {self.slow_query_threshold}s)"
             )
 
+    async def execute_batch_sqls_for_mcp(
+            self, sqls: list[str],
+            timeout: int = 30,
+            session_id: str = "mcp_session",
+            user_id: str = "mcp_user",
+            auth_context=None
+    ) -> dict[str, Any]:
+        """Execute multiple sqls in batch"""
+        if not sqls:
+            return {
+                "success": False,
+                "error": "SQL query is required",
+                "data": None
+            }
+        query_requests = [
+            QueryRequest(
+                sql=sql,
+                session_id=session_id,
+                user_id=user_id,
+                timeout=timeout,
+                cache_enabled=False
+            )
+            for sql in sqls
+        ]
+        query_results = await self.execute_batch_queries(query_requests, auth_context)
+        # Serialize data for JSON response
+        results = [
+            {
+                "data": [self._serialize_row_data(data) for data in result.data],
+                "row_count": result.row_count,
+                "execution_time": result.execution_time,
+                "metadata": {
+                    "columns": result.metadata.get("columns", []),
+                    "query": result.sql
+                }
+            }
+            for result in query_results
+        ]
+
+        return {
+            "success": True,
+            "multiple_results": True,
+            "results": results
+        }
+
     async def execute_batch_queries(
         self, query_requests: list[QueryRequest], auth_context=None
     ) -> list[QueryResult]:
@@ -484,20 +532,24 @@ class DorisQueryExecutor:
             self.execute_query(request, auth_context) for request in query_requests
         ]
 
-        try:
+        query_results = []
         results = await asyncio.gather(*tasks, return_exceptions=True)
-        except Exception as e:
-            self.logger.error(f"Batch query execution failed: {e}")
-            raise
+        for result in results:
+            if isinstance(result, Exception):
+                self.logger.error(f"Batch query execution failed: {result}")
+                raise result
+            else:
+                query_results.append(result)
 
-        return results
+        return query_results
 
     async def explain_query(self, sql: str, session_id: str) -> dict[str, Any]:
         """Get query execution plan"""
         explain_sql = f"EXPLAIN {sql}"
 
         connection = await self.connection_manager.get_connection(session_id)
-        result = await connection.execute(explain_sql)
+        auth_context = get_auth_context()
+        result = await connection.execute(explain_sql, auth_context=auth_context)
 
         return {
             "query": sql,
@@ -546,9 +598,13 @@ class DorisQueryExecutor:
         limit: int = 1000,
         timeout: int = 30,
         session_id: str = "mcp_session",
-        user_id: str = "mcp_user"
+        user_id: str = "mcp_user",
+        auth_context = None  # FIX for Issue #62 Bug 1: Accept auth_context with token
     ) -> Dict[str, Any]:
-        """Execute SQL query for MCP interface - unified method"""
+        """Execute SQL query for MCP interface - unified method
+
+        FIX for Issue #62 Bug 1: Now accepts auth_context parameter to support token-bound database configuration
+        """
         max_retries = 2
         retry_count = 0
 
@@ -564,19 +620,31 @@ class DorisQueryExecutor:
                 # Import required security modules
                 from .security import DorisSecurityManager, AuthContext, SecurityLevel
 
-                # Create proper auth context with read-only permissions
+                # FIX: Use provided auth_context if available (contains token for DB config)
+                # Otherwise create default auth context for backward compatibility
+                if auth_context is None:
                     auth_context = AuthContext(
                         user_id=user_id,
                         roles=["read_only_user"],  # Restrictive role for MCP interface
                         permissions=["read_data"],  # Only read permissions
                         session_id=session_id,
-                    security_level=SecurityLevel.INTERNAL
+                        security_level=SecurityLevel.INTERNAL,
+                        token=""  # No token in default context
                     )
+                else:
+                    # Use provided auth_context (may contain token for database configuration)
+                    self.logger.debug(f"Using provided auth_context with token: {bool(hasattr(auth_context, 'token') and auth_context.token)}")
 
                 # Perform SQL security validation if enabled
                 if hasattr(self.connection_manager, 'config') and hasattr(self.connection_manager.config, 'security'):
                     if self.connection_manager.config.security.enable_security_check:
                         try:
+                            # 🔧 FIX: Use existing security_manager to avoid creating multiple TokenManager instances
+                            # Creating new DorisSecurityManager each time causes multiple hot reload monitors
+                            security_manager = getattr(self.connection_manager, 'security_manager', None)
+                            if not security_manager:
+                                # Fallback: create new one only if not available (should rarely happen)
+                                self.logger.warning("No existing security_manager, creating new instance")
                                 security_manager = DorisSecurityManager(self.connection_manager.config)
                             validation_result = await security_manager.validate_sql_security(sql, auth_context)
 
@@ -623,6 +691,15 @@ class DorisQueryExecutor:
                         sql = sql[:-1]
                     sql = f"{sql} LIMIT {limit}"
                 
+                all_statements = [
+                    s.strip()
+                    for s in sqlparse.split(sql)
+                    if s.strip()
+                ]
+                if len(all_statements) > 1:
+                    return await self.execute_batch_sqls_for_mcp(sqls=all_statements, timeout=timeout,
+                                                                 session_id=session_id, user_id=user_id,
+                                                                 auth_context=auth_context)
                 # Create query request
                 query_request = QueryRequest(
                     sql=sql,
@@ -880,17 +957,20 @@ async def execute_sql_query(sql: str, connection_manager: DorisConnectionManager
 
     This function now includes security validation to ensure safe query execution.
     All queries are validated against the configured security policies before execution.
+
+    FIX for Issue #62 Bug 1: Now supports auth_context parameter for token-bound database configuration
+    FIX for Issue #58 Problem 2: Removed executor.close() to prevent ClosedResourceError in multi-worker mode
     """
     try:
         # Create query executor with the connection manager's configuration
         executor = DorisQueryExecutor(connection_manager)
 
-        try:
         # Extract parameters from kwargs or use defaults
         limit = kwargs.get("limit", 1000)
         timeout = kwargs.get("timeout", 30)
         session_id = kwargs.get("session_id", "mcp_session")
         user_id = kwargs.get("user_id", "mcp_user")
+        auth_context = kwargs.get("auth_context", None)  # FIX: Extract auth_context
 
         # The execute_sql_for_mcp method now includes security validation
         result = await executor.execute_sql_for_mcp(
@@ -898,11 +978,17 @@ async def execute_sql_query(sql: str, connection_manager: DorisConnectionManager
             limit=limit,
             timeout=timeout,
             session_id=session_id,
-                user_id=user_id
+            user_id=user_id,
+            auth_context=auth_context  # FIX: Pass auth_context with token
         )
+
+        # FIX for Issue #58 Problem 2: Do NOT close executor here
+        # In multi-worker mode, closing here causes ClosedResourceError
+        # The executor's resources (cache, background tasks) will be managed
+        # by the connection_manager lifecycle and Python's garbage collection
+        # This prevents premature cleanup while MCP session manager is still processing
+
         return result
-        finally:
-            await executor.close()
 
     except Exception as e:
         return {

doris_mcp_server/utils/schema_extractor.py

@@ -32,6 +32,11 @@ from datetime import datetime, timedelta
 
 # Import unified logging configuration
 from .logger import get_logger
+from .sql_security_utils import (
+    SQLSecurityError,
+    validate_identifier,
+    quote_identifier
+)
 
 # Configure logging
 logger = get_logger(__name__)
@@ -431,6 +436,16 @@ class MetadataExtractor:
             logger.warning("Database name not specified")
             return {}
         
+        # SECURITY FIX: Validate identifiers to prevent SQL injection
+        try:
+            validate_identifier(table_name, "table name")
+            validate_identifier(db_name, "database name")
+            if effective_catalog:
+                validate_identifier(effective_catalog, "catalog name")
+        except SQLSecurityError as e:
+            logger.warning(f"Invalid identifier rejected in get_table_schema: {e}")
+            return {}
+        
         cache_key = f"schema_{effective_catalog or 'default'}_{db_name}_{table_name}"
         if cache_key in self.metadata_cache and (datetime.now() - self.metadata_cache_time.get(cache_key, datetime.min)).total_seconds() < self.cache_ttl:
             return self.metadata_cache[cache_key]
@@ -536,6 +551,16 @@ class MetadataExtractor:
             logger.warning("Database name not specified")
             return ""
         
+        # SECURITY FIX: Validate identifiers
+        try:
+            validate_identifier(table_name, "table name")
+            validate_identifier(db_name, "database name")
+            if effective_catalog:
+                validate_identifier(effective_catalog, "catalog name")
+        except SQLSecurityError as e:
+            logger.warning(f"Invalid identifier rejected: {e}")
+            return ""
+        
         cache_key = f"table_comment_{effective_catalog or 'default'}_{db_name}_{table_name}"
         if cache_key in self.metadata_cache and (datetime.now() - self.metadata_cache_time.get(cache_key, datetime.min)).total_seconds() < self.cache_ttl:
             return self.metadata_cache[cache_key]
@@ -587,6 +612,16 @@ class MetadataExtractor:
             logger.warning("Database name not specified")
             return {}
         
+        # SECURITY FIX: Validate identifiers
+        try:
+            validate_identifier(table_name, "table name")
+            validate_identifier(db_name, "database name")
+            if effective_catalog:
+                validate_identifier(effective_catalog, "catalog name")
+        except SQLSecurityError as e:
+            logger.warning(f"Invalid identifier rejected: {e}")
+            return {}
+        
         cache_key = f"column_comments_{effective_catalog or 'default'}_{db_name}_{table_name}"
         if cache_key in self.metadata_cache and (datetime.now() - self.metadata_cache_time.get(cache_key, datetime.min)).total_seconds() < self.cache_ttl:
             return self.metadata_cache[cache_key]
@@ -643,17 +678,30 @@ class MetadataExtractor:
             logger.error("Database name not specified")
             return []
         
+        # SECURITY FIX: Validate identifiers
+        try:
+            validate_identifier(table_name, "table name")
+            validate_identifier(db_name, "database name")
+            if effective_catalog:
+                validate_identifier(effective_catalog, "catalog name")
+        except SQLSecurityError as e:
+            logger.warning(f"Invalid identifier rejected: {e}")
+            return []
+        
         cache_key = f"indexes_{effective_catalog or 'default'}_{db_name}_{table_name}"
         if cache_key in self.metadata_cache and (datetime.now() - self.metadata_cache_time.get(cache_key, datetime.min)).total_seconds() < self.cache_ttl:
             return self.metadata_cache[cache_key]
         
         try:
-            # Build query with catalog prefix if specified
+            # Build query with catalog prefix if specified (identifiers already validated)
+            safe_table = quote_identifier(table_name, "table name")
+            safe_db = quote_identifier(db_name, "database name")
             if effective_catalog:
-                query = f"SHOW INDEX FROM `{effective_catalog}`.`{db_name}`.`{table_name}`"
+                safe_catalog = quote_identifier(effective_catalog, "catalog name")
+                query = f"SHOW INDEX FROM {safe_catalog}.{safe_db}.{safe_table}"
                 logger.info(f"Using three-part naming for index query: {query}")
             else:
-                query = f"SHOW INDEX FROM `{db_name}`.`{table_name}`"
+                query = f"SHOW INDEX FROM {safe_db}.{safe_table}"
             
             try:
                 # NOTE: Deprecated sync path retained for compatibility; use async variant instead.
@@ -1146,8 +1194,17 @@ class MetadataExtractor:
         """
         try:
             if self.connection_manager:
+                # FIX: Get auth_context from global ContextVar for token-bound database configuration
+                # This ensures all query methods use the correct user's connection pool
+                auth_context = None
+                try:
+                    from .security import mcp_auth_context_var
+                    auth_context = mcp_auth_context_var.get()
+                except Exception:
+                    pass
+                
                 # Use the injected connection manager directly (async)
-                result = await self.connection_manager.execute_query(self._session_id, query, None)
+                result = await self.connection_manager.execute_query(self._session_id, query, None, auth_context)
                 
                 # Extract data from QueryResult
                 if hasattr(result, 'data'):
@@ -1188,12 +1245,28 @@ class MetadataExtractor:
         try:
             # Use async query method
             effective_catalog = catalog_name or self.catalog_name
+            effective_db = db_name or self.db_name
             
-            # Build query statement
+            # SECURITY FIX: Validate identifiers
+            try:
+                validate_identifier(table_name, "table name")
+                if effective_db:
+                    validate_identifier(effective_db, "database name")
                 if effective_catalog and effective_catalog != "internal":
-                query = f"DESCRIBE `{effective_catalog}`.`{db_name or self.db_name}`.`{table_name}`"
+                    validate_identifier(effective_catalog, "catalog name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            # Build query statement using safe identifiers
+            safe_table = quote_identifier(table_name, "table name")
+            safe_db = quote_identifier(effective_db, "database name") if effective_db else None
+            
+            if effective_catalog and effective_catalog != "internal":
+                safe_catalog = quote_identifier(effective_catalog, "catalog name")
+                query = f"DESCRIBE {safe_catalog}.{safe_db}.{safe_table}"
             else:
-                query = f"DESCRIBE `{db_name or self.db_name}`.`{table_name}`"
+                query = f"DESCRIBE {safe_db}.{safe_table}"
             
             # Execute async query
             result = await self._execute_query_async(query, db_name)
@@ -1226,8 +1299,15 @@ class MetadataExtractor:
         try:
             effective_catalog = catalog_name or self.catalog_name
             
+            # SECURITY FIX: Validate catalog name if provided
             if effective_catalog and effective_catalog != "internal":
-                query = f"SHOW DATABASES FROM `{effective_catalog}`"
+                try:
+                    validate_identifier(effective_catalog, "catalog name")
+                except SQLSecurityError as e:
+                    logger.warning(f"Invalid catalog name rejected: {e}")
+                    return []
+                safe_catalog = quote_identifier(effective_catalog, "catalog name")
+                query = f"SHOW DATABASES FROM {safe_catalog}"
             else:
                 query = "SHOW DATABASES"
             
@@ -1257,10 +1337,23 @@ class MetadataExtractor:
             effective_catalog = catalog_name or self.catalog_name
             effective_db = db_name or self.db_name
             
+            # SECURITY FIX: Validate identifiers
+            try:
+                if effective_db:
+                    validate_identifier(effective_db, "database name")
                 if effective_catalog and effective_catalog != "internal":
-                query = f"SHOW TABLES FROM `{effective_catalog}`.`{effective_db}`"
+                    validate_identifier(effective_catalog, "catalog name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+            
+            safe_db = quote_identifier(effective_db, "database name") if effective_db else None
+            
+            if effective_catalog and effective_catalog != "internal":
+                safe_catalog = quote_identifier(effective_catalog, "catalog name")
+                query = f"SHOW TABLES FROM {safe_catalog}.{safe_db}"
             else:
-                query = f"SHOW TABLES FROM `{effective_db}`"
+                query = f"SHOW TABLES FROM {safe_db}"
             
             result = await self._execute_query_async(query, effective_db)
             
@@ -1319,6 +1412,15 @@ class MetadataExtractor:
             effective_db = db_name or self.db_name
             effective_catalog = catalog_name or self.catalog_name
 
+            # SECURITY FIX: Validate identifiers
+            try:
+                validate_identifier(table_name, "table name")
+                if effective_db:
+                    validate_identifier(effective_db, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return ""
+
             query = f"""
             SELECT 
                 TABLE_COMMENT 
@@ -1343,6 +1445,15 @@ class MetadataExtractor:
             effective_db = db_name or self.db_name
             effective_catalog = catalog_name or self.catalog_name
 
+            # SECURITY FIX: Validate identifiers
+            try:
+                validate_identifier(table_name, "table name")
+                if effective_db:
+                    validate_identifier(effective_db, "database name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return {}
+
             query = f"""
             SELECT 
                 COLUMN_NAME, 
@@ -1373,12 +1484,27 @@ class MetadataExtractor:
             effective_db = db_name or self.db_name
             effective_catalog = catalog_name or self.catalog_name
 
-            # Build query with catalog prefix if specified
+            # SECURITY FIX: Validate identifiers
+            try:
+                validate_identifier(table_name, "table name")
+                if effective_db:
+                    validate_identifier(effective_db, "database name")
                 if effective_catalog:
-                query = f"SHOW INDEX FROM `{effective_catalog}`.`{effective_db}`.`{table_name}`"
+                    validate_identifier(effective_catalog, "catalog name")
+            except SQLSecurityError as e:
+                logger.warning(f"Invalid identifier rejected: {e}")
+                return []
+
+            # Build query with catalog prefix if specified (using safe identifiers)
+            safe_table = quote_identifier(table_name, "table name")
+            safe_db = quote_identifier(effective_db, "database name") if effective_db else None
+            
+            if effective_catalog:
+                safe_catalog = quote_identifier(effective_catalog, "catalog name")
+                query = f"SHOW INDEX FROM {safe_catalog}.{safe_db}.{safe_table}"
                 logger.info(f"Using three-part naming for async index query: {query}")
             else:
-                query = f"SHOW INDEX FROM `{effective_db}`.`{table_name}`"
+                query = f"SHOW INDEX FROM {safe_db}.{safe_table}"
 
             rows = await self._execute_query_async(query, effective_db)
             indexes: List[Dict[str, Any]] = []
@@ -1464,6 +1590,9 @@ class MetadataExtractor:
         """
         Execute SQL query and return results, supports catalog federation queries
         Unified interface for MCP tools
+
+        FIX for Issue #62 Bug 1: Now retrieves auth_context from context variable to support token-bound database configuration
+        FIX for Issue #62 Bug 3: Now uses db_name and catalog_name parameters to switch database context
         """
         logger.info(f"Executing SQL query: {sql}, DB: {db_name}, Catalog: {catalog_name}, MaxRows: {max_rows}, Timeout: {timeout}")
 
@@ -1471,15 +1600,86 @@ class MetadataExtractor:
             if not sql:
                 return self._format_response(success=False, error="No SQL statement provided", message="Please provide SQL statement to execute")
 
+            # FIX for Issue #62 Bug 3: Build context switching SQL if db_name or catalog_name is specified
+            # SECURITY FIX: Validate catalog_name and db_name to prevent SQL injection
+            final_sql = sql
+            if catalog_name or db_name:
+                context_statements = []
+
+                # Validate and sanitize catalog_name
+                if catalog_name:
+                    try:
+                        validate_identifier(catalog_name, "catalog name")
+                    except SQLSecurityError as e:
+                        logger.warning(f"Invalid catalog name rejected: {e}")
+                        return self._format_response(
+                            success=False, 
+                            error=f"Invalid catalog name: {catalog_name}", 
+                            message="Catalog name contains invalid characters"
+                        )
+                    # Use quote_identifier to safely escape the catalog name
+                    safe_catalog = quote_identifier(catalog_name, "catalog name")
+                    context_statements.append(f"USE CATALOG {safe_catalog}")
+                    logger.debug(f"Switching to catalog: {catalog_name}")
+
+                # Validate and sanitize db_name
+                if db_name:
+                    try:
+                        validate_identifier(db_name, "database name")
+                    except SQLSecurityError as e:
+                        logger.warning(f"Invalid database name rejected: {e}")
+                        return self._format_response(
+                            success=False, 
+                            error=f"Invalid database name: {db_name}", 
+                            message="Database name contains invalid characters"
+                        )
+                    # Use quote_identifier to safely escape the database name
+                    safe_db = quote_identifier(db_name, "database name")
+                    if catalog_name:
+                        safe_catalog = quote_identifier(catalog_name, "catalog name")
+                        context_statements.append(f"USE {safe_catalog}.{safe_db}")
+                    else:
+                        context_statements.append(f"USE {safe_db}")
+                    logger.debug(f"Switching to database: {db_name}")
+
+                # Combine context switching with original SQL
+                if context_statements:
+                    # Remove trailing semicolon from context statements if present
+                    context_sql = "; ".join(context_statements)
+                    # Ensure original SQL doesn't start with semicolon
+                    sql_clean = sql.lstrip(";").strip()
+                    final_sql = f"{context_sql}; {sql_clean}"
+                    logger.debug(f"Modified SQL with context switching: {final_sql[:200]}...")
+
+            # FIX: Try to get auth_context from context variable (set by HTTP middleware)
+            # This allows token-bound database configuration to work
+            # CRITICAL: Use the global ContextVar from security.py to ensure same instance is used everywhere
+            auth_context = None
+            try:
+                from .security import mcp_auth_context_var
+
+                # Get auth_context from the global context variable
+                # This will be set by the HTTP request handler in main.py
+                auth_context = mcp_auth_context_var.get()
+
+                if auth_context:
+                    logger.debug(f"Retrieved auth_context from context variable with token: {bool(hasattr(auth_context, 'token') and auth_context.token)}")
+                else:
+                    logger.debug("No auth_context found in context variable, using default")
+            except Exception as ctx_error:
+                logger.debug(f"Could not retrieve auth_context from context variable: {ctx_error}")
+                auth_context = None
+
             # Import query executor
             from .query_executor import execute_sql_query
 
-            # Call execute_sql_query to execute query
+            # Call execute_sql_query to execute query with auth_context
             exec_result = await execute_sql_query(
-                sql=sql,
+                sql=final_sql,  # Use modified SQL with context switching
                 connection_manager=self.connection_manager,
                 limit=max_rows,
-                timeout=timeout
+                timeout=timeout,
+                auth_context=auth_context  # FIX: Pass auth_context with token
             )
 
             return exec_result
@@ -1500,6 +1700,36 @@ class MetadataExtractor:
         if not table_name:
             return self._format_response(success=False, error="Missing table_name parameter")
         
+        # SECURITY: Validate identifiers before processing
+        try:
+            validate_identifier(table_name, "table name")
+        except SQLSecurityError as e:
+            return self._format_response(
+                success=False, 
+                error=f"Invalid table name: {table_name}",
+                message="Table name contains invalid characters"
+            )
+        
+        if db_name:
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid database name: {db_name}",
+                    message="Database name contains invalid characters"
+                )
+        
+        if catalog_name and catalog_name != "internal":
+            try:
+                validate_identifier(catalog_name, "catalog name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid catalog name: {catalog_name}",
+                    message="Catalog name contains invalid characters"
+                )
+        
         try:
             schema = await self.get_table_schema_async(table_name=table_name, db_name=db_name, catalog_name=catalog_name)
             
@@ -1523,6 +1753,27 @@ class MetadataExtractor:
         """Get list of all table names in specified database - MCP interface"""
         logger.info(f"Getting database table list: DB: {db_name}, Catalog: {catalog_name}")
         
+        # SECURITY: Validate identifiers
+        if db_name:
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid database name: {db_name}",
+                    message="Database name contains invalid characters"
+                )
+        
+        if catalog_name and catalog_name != "internal":
+            try:
+                validate_identifier(catalog_name, "catalog name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid catalog name: {catalog_name}",
+                    message="Catalog name contains invalid characters"
+                )
+        
         try:
             tables = await self.get_database_tables_async(db_name=db_name, catalog_name=catalog_name)
             return self._format_response(success=True, result=tables)
@@ -1553,6 +1804,36 @@ class MetadataExtractor:
         if not table_name:
             return self._format_response(success=False, error="Missing table_name parameter")
         
+        # SECURITY: Validate identifiers
+        try:
+            validate_identifier(table_name, "table name")
+        except SQLSecurityError as e:
+            return self._format_response(
+                success=False,
+                error=f"Invalid table name: {table_name}",
+                message="Table name contains invalid characters"
+            )
+        
+        if db_name:
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid database name: {db_name}",
+                    message="Database name contains invalid characters"
+                )
+        
+        if catalog_name and catalog_name != "internal":
+            try:
+                validate_identifier(catalog_name, "catalog name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid catalog name: {catalog_name}",
+                    message="Catalog name contains invalid characters"
+                )
+        
         try:
             comment = await self.get_table_comment_async(table_name=table_name, db_name=db_name, catalog_name=catalog_name)
             return self._format_response(success=True, result=comment)
@@ -1572,6 +1853,36 @@ class MetadataExtractor:
         if not table_name:
             return self._format_response(success=False, error="Missing table_name parameter")
         
+        # SECURITY: Validate identifiers
+        try:
+            validate_identifier(table_name, "table name")
+        except SQLSecurityError as e:
+            return self._format_response(
+                success=False,
+                error=f"Invalid table name: {table_name}",
+                message="Table name contains invalid characters"
+            )
+        
+        if db_name:
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid database name: {db_name}",
+                    message="Database name contains invalid characters"
+                )
+        
+        if catalog_name and catalog_name != "internal":
+            try:
+                validate_identifier(catalog_name, "catalog name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid catalog name: {catalog_name}",
+                    message="Catalog name contains invalid characters"
+                )
+        
         try:
             comments = await self.get_column_comments_async(table_name=table_name, db_name=db_name, catalog_name=catalog_name)
             return self._format_response(success=True, result=comments)
@@ -1591,6 +1902,36 @@ class MetadataExtractor:
         if not table_name:
             return self._format_response(success=False, error="Missing table_name parameter")
         
+        # SECURITY: Validate identifiers
+        try:
+            validate_identifier(table_name, "table name")
+        except SQLSecurityError as e:
+            return self._format_response(
+                success=False,
+                error=f"Invalid table name: {table_name}",
+                message="Table name contains invalid characters"
+            )
+        
+        if db_name:
+            try:
+                validate_identifier(db_name, "database name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid database name: {db_name}",
+                    message="Database name contains invalid characters"
+                )
+        
+        if catalog_name and catalog_name != "internal":
+            try:
+                validate_identifier(catalog_name, "catalog name")
+            except SQLSecurityError as e:
+                return self._format_response(
+                    success=False,
+                    error=f"Invalid catalog name: {catalog_name}",
+                    message="Catalog name contains invalid characters"
+                )
+        
         try:
             indexes = await self.get_table_indexes_async(table_name=table_name, db_name=db_name, catalog_name=catalog_name)
             return self._format_response(success=True, result=indexes)

doris_mcp_server/utils/security.py

@@ -22,6 +22,7 @@ Implements enterprise-level authentication, authorization, SQL security validati
 
 import logging
 import re
+from contextvars import ContextVar
 from dataclasses import dataclass, field
 from datetime import datetime
 from enum import Enum
@@ -34,6 +35,10 @@ from sqlparse.tokens import Keyword, Name
 from .logger import get_logger
 from .config import DatabaseConfig
 
+# Global ContextVar for auth_context - must be a single instance shared across all modules
+# This allows token-bound database configuration to work correctly in concurrent requests
+mcp_auth_context_var: ContextVar['AuthContext'] = ContextVar('mcp_auth_context', default=None)
+
 
 class SecurityLevel(Enum):
     """Security level enumeration"""
@@ -903,27 +908,47 @@ class SQLSecurityValidator:
             return ValidationResult(is_valid=True)
 
         try:
-            # Parse SQL statement
-            parsed = sqlparse.parse(sql)[0]
+            # SECURITY FIX: Parse ALL SQL statements, not just the first one
+            # This prevents bypassing security checks by injecting additional statements
+            all_statements = sqlparse.parse(sql)
+
+            if not all_statements:
+                return ValidationResult(
+                    is_valid=False,
+                    error_message="Empty or invalid SQL statement",
+                    risk_level="medium"
+                )
+
+            # SECURITY FIX: Validate each statement individually
+            for idx, parsed in enumerate(all_statements):
+                # Skip empty statements (e.g., from trailing semicolons)
+                if not parsed.tokens or str(parsed).strip() == '':
+                    continue
+
+                self.logger.debug(f"Validating SQL statement {idx + 1}/{len(all_statements)}: {str(parsed)[:100]}...")
 
                 # Check blocked operations first (more specific)
                 keyword_result = await self._check_blocked_keywords(parsed)
                 if not keyword_result.is_valid:
+                    keyword_result.error_message = f"Statement {idx + 1}: {keyword_result.error_message}"
                     return keyword_result
 
                 # Check SQL injection risks
                 injection_result = await self._check_sql_injection(sql, parsed)
                 if not injection_result.is_valid:
+                    injection_result.error_message = f"Statement {idx + 1}: {injection_result.error_message}"
                     return injection_result
 
                 # Check query complexity
                 complexity_result = await self._check_query_complexity(parsed)
                 if not complexity_result.is_valid:
+                    complexity_result.error_message = f"Statement {idx + 1}: {complexity_result.error_message}"
                     return complexity_result
 
                 # Check table access permissions
                 table_result = await self._check_table_access(parsed, auth_context)
                 if not table_result.is_valid:
+                    table_result.error_message = f"Statement {idx + 1}: {table_result.error_message}"
                     return table_result
 
             return ValidationResult(is_valid=True)
@@ -939,28 +964,69 @@ class SQLSecurityValidator:
     async def _check_sql_injection(
         self, sql: str, parsed: Statement
     ) -> ValidationResult:
-        """Check SQL injection risks"""
-        # Check common SQL injection patterns
+        """Check SQL injection risks with improved pattern detection
+
+        FIX for Issue #62 Bug 2: Improved patterns to reduce false positives
+        Now better distinguishes between legitimate SQL (like BETWEEN...AND) and injection attempts
+        """
+        # Improved injection patterns that are more specific and less prone to false positives
         injection_patterns = [
-            r"(?i)(?<![A-Za-z0-9_])(union|select|insert|update|delete|drop|create|alter)(?![A-Za-z0-9_])\s+[\s\S]*?\s+(?<![A-Za-z0-9_])(union|select|insert|update|delete|drop|create|alter)(?![A-Za-z0-9_])",
-            r"(\s|^)(or|and)\s+\d+\s*=\s*\d+",
-            r"(\s|^)(or|and)\s+['\"].*['\"]",
-            r";\s*(drop|delete|truncate|alter|create)",
-            r"(exec|execute|sp_|xp_)",
-            r"(script|javascript|vbscript)",
-            r"(char|ascii|substring|concat)\s*\(",
+            # Stacked queries with dangerous operations (true injection risk)
+            r";\s*(DROP|DELETE|TRUNCATE|ALTER|CREATE|INSERT|UPDATE)\s+",
+
+            # UNION-based injection (but allow legitimate UNION queries)
+            # Only flag if UNION is followed by suspicious patterns like SELECT with WHERE 1=1
+            r"UNION\s+(ALL\s+)?SELECT\s+.*\s+(WHERE|AND|OR)\s+\d+\s*=\s*\d+",
+
+            # Boolean-based blind injection with comments (true injection pattern)
+            r"(WHERE|AND|OR)\s+\d+\s*=\s*\d+\s*(--|#|/\*)",
+
+            # Quote-based injection attempts (but not in legitimate strings)
+            r"(WHERE|AND|OR)\s+(['\"])[^\2]*\2\s*=\s*\2[^\2]*\2",
+
+            # Time-based blind injection
+            r"(SLEEP|WAITFOR|BENCHMARK)\s*\(",
+
+            # System stored procedure injection
+            r"(EXEC|EXECUTE|SP_|XP_)\s*\(",
+
+            # Script injection attempts
+            r"<\s*(SCRIPT|JAVASCRIPT|VBSCRIPT)",
         ]
 
-        sql_lower = sql.lower()
+        # FIX: Don't flag legitimate SQL functions and keywords
+        # These patterns are too broad and cause false positives:
+        # - REMOVED: r"(char|ascii|substring|concat)\s*\(" - These are legitimate SQL functions
+        # - REMOVED: r"(\s|^)(or|and)\s+\d+\s*=\s*\d+" - This flags BETWEEN...AND constructs
+        # - REMOVED: r"(\s|^)(or|and)\s+['\"].*['\"]" - This is too broad
+
+        sql_upper = sql.upper()
+
+        # Special case: Allow BETWEEN...AND which is legitimate SQL
+        # This prevents false positives like "WHERE dt BETWEEN '2025-01-01' AND '2025-01-31'"
+        if "BETWEEN" in sql_upper and "AND" in sql_upper:
+            # This is likely a BETWEEN clause, not injection
+            # Check if AND appears in a BETWEEN context
+            between_pattern = r"BETWEEN\s+[^\s]+\s+AND\s+[^\s]+"
+            if re.search(between_pattern, sql_upper, re.IGNORECASE):
+                # Remove BETWEEN clauses before checking other patterns
+                sql_cleaned = re.sub(between_pattern, "BETWEEN_CLAUSE", sql_upper, flags=re.IGNORECASE)
+                sql_to_check = sql_cleaned
+            else:
+                sql_to_check = sql_upper
+        else:
+            sql_to_check = sql_upper
+
         for pattern in injection_patterns:
-            if re.search(pattern, sql_lower, re.IGNORECASE):
+            if re.search(pattern, sql_to_check, re.IGNORECASE):
+                self.logger.warning(f"Potential SQL injection pattern detected: {pattern}")
                 return ValidationResult(
                     is_valid=False,
                     error_message="Potential SQL injection risk detected",
                     risk_level="high",
                 )
 
-        # Check suspicious quotes and comments
+        # Check suspicious quotes and comments (with improved detection)
         if self._has_suspicious_quotes_or_comments(sql):
             return ValidationResult(
                 is_valid=False,
@@ -971,20 +1037,68 @@ class SQLSecurityValidator:
         return ValidationResult(is_valid=True)
 
     def _has_suspicious_quotes_or_comments(self, sql: str) -> bool:
-        """Check suspicious quote and comment patterns"""
-        # Check unmatched quotes
-        single_quotes = sql.count("'")
-        double_quotes = sql.count('"')
+        """Check suspicious quote and comment patterns with improved detection
 
+        FIX for Issue #62 Bug 2: Improved detection to reduce false positives
+        Now distinguishes between legitimate comments/strings and injection attempts
+        """
+        try:
+            # Use sqlparse to parse the SQL and distinguish between code and comments/strings
+            import sqlparse
+            from sqlparse.tokens import Comment, String
+
+            # Parse the SQL
+            parsed = sqlparse.parse(sql)
+            if not parsed:
+                # If parsing fails, be conservative
+                return True
+
+            statement = parsed[0]
+
+            # Check for unmatched quotes ONLY in non-string tokens
+            # This prevents false positives from legitimate string content
+            non_string_content = []
+            has_string_tokens = False
+
+            for token in statement.flatten():
+                if token.ttype in (String.Single, String.Double):
+                    has_string_tokens = True
+                    # Skip string content - quotes inside strings are legitimate
+                    continue
+                elif token.ttype in (Comment.Single, Comment.Multi):
+                    # Comments are generally OK, but check for suspicious injection patterns
+                    comment_value = str(token).lower()
+                    # Check if comment contains dangerous SQL keywords
+                    dangerous_in_comments = ['drop', 'delete', 'insert', 'update', 'union', 'exec', 'execute']
+                    if any(keyword in comment_value for keyword in dangerous_in_comments):
+                        self.logger.warning(f"Suspicious SQL keyword in comment: {token}")
+                        return True
+                    # Normal comments are OK
+                    continue
+                else:
+                    # Accumulate non-string, non-comment content
+                    non_string_content.append(str(token))
+
+            # Check for unmatched quotes in non-string content
+            non_string_text = ''.join(non_string_content)
+            single_quotes = non_string_text.count("'")
+            double_quotes = non_string_text.count('"')
+
+            # Only flag if there are unmatched quotes in actual SQL code (not in strings)
             if single_quotes % 2 != 0 or double_quotes % 2 != 0:
                 return True
 
-        # Check SQL comments
-        if "--" in sql or "/*" in sql:
-            return True
+            # FIX: Don't flag legitimate SQL comments
+            # Comments are OK as long as they don't contain dangerous patterns (already checked above)
 
             return False
 
+        except Exception as e:
+            self.logger.debug(f"SQL parsing error in quote/comment check: {e}")
+            # On parsing error, fall back to conservative check
+            # But be more lenient than before
+            return False  # Don't flag on parse errors to reduce false positives
+
     async def _check_blocked_keywords(self, parsed: Statement) -> ValidationResult:
         """Check blocked keywords"""
         blocked_operations = []
@@ -1045,6 +1159,10 @@ class SQLSecurityValidator:
         self, parsed: Statement, auth_context: AuthContext
     ) -> ValidationResult:
         """Check table access permissions"""
+        # If no auth_context, skip table access checks (rely on other security checks)
+        if auth_context is None:
+            return ValidationResult(is_valid=True)
+        
         # Extract table names from query
         tables = self._extract_table_names(parsed)
 

doris_mcp_server/utils/security_analytics_tools.py

@@ -26,6 +26,7 @@ from collections import Counter, defaultdict
 
 from .db import DorisConnectionManager
 from .logger import get_logger
+from .sql_security_utils import get_auth_context
 
 logger = get_logger(__name__)
 
@@ -192,7 +193,9 @@ class SecurityAnalyticsTools:
             LIMIT 10000
             """
             
-            result = await connection.execute(audit_sql)
+            # SECURITY FIX: Pass auth_context to execute
+            auth_context = get_auth_context()
+            result = await connection.execute(audit_sql, auth_context=auth_context)
             return result.data if result.data else []
             
         except Exception as e:
@@ -215,7 +218,8 @@ class SecurityAnalyticsTools:
                 LIMIT 10000
                 """
                 
-                result = await connection.execute(simple_audit_sql)
+                auth_context = get_auth_context()
+                result = await connection.execute(simple_audit_sql, auth_context=auth_context)
                 return result.data if result.data else []
                 
             except Exception as e2:
@@ -498,7 +502,8 @@ class SecurityAnalyticsTools:
             FROM mysql.user
             """
             
-            result = await connection.execute(roles_sql)
+            auth_context = get_auth_context()
+            result = await connection.execute(roles_sql, auth_context=auth_context)
             
             user_roles = defaultdict(list)
             if result.data:

doris_mcp_server/utils/sql_security_utils.py

@@ -0,0 +1,301 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+"""
+SQL Security Utilities Module
+
+Provides SQL identifier validation, escaping, and safe query building utilities
+to prevent SQL injection attacks.
+"""
+
+import re
+from contextvars import ContextVar
+from typing import Optional, Tuple, List, Any
+
+from .logger import get_logger
+
+logger = get_logger(__name__)
+
+# Context variable for auth_context (set by HTTP middleware)
+auth_context_var: ContextVar = ContextVar('mcp_auth_context', default=None)
+
+
+class SQLSecurityError(Exception):
+    """Exception raised for SQL security validation failures"""
+    pass
+
+
+class SQLSecurityUtils:
+    """
+    SQL Security Utilities for preventing SQL injection attacks.
+    
+    Provides:
+    - Identifier validation (database names, table names, column names)
+    - Safe identifier quoting with backticks
+    - Safe table reference building
+    - Auth context retrieval from context variables
+    """
+    
+    # Valid SQL identifier pattern: letters, numbers, underscores
+    # Must start with letter or underscore, not a number
+    # Supports Unicode letters for international database/table names
+    IDENTIFIER_PATTERN = re.compile(r'^[a-zA-Z_\u4e00-\u9fff][a-zA-Z0-9_\u4e00-\u9fff]*$')
+    
+    # Maximum identifier length (MySQL/Doris standard)
+    MAX_IDENTIFIER_LENGTH = 64
+    
+    # SQL reserved keywords that should be quoted
+    SQL_KEYWORDS = {
+        'SELECT', 'FROM', 'WHERE', 'INSERT', 'UPDATE', 'DELETE', 'DROP', 
+        'CREATE', 'ALTER', 'TABLE', 'DATABASE', 'INDEX', 'VIEW', 'AND', 
+        'OR', 'NOT', 'NULL', 'TRUE', 'FALSE', 'IN', 'LIKE', 'BETWEEN',
+        'JOIN', 'LEFT', 'RIGHT', 'INNER', 'OUTER', 'ON', 'AS', 'ORDER',
+        'BY', 'GROUP', 'HAVING', 'LIMIT', 'OFFSET', 'UNION', 'ALL',
+        'DISTINCT', 'INTO', 'VALUES', 'SET', 'DEFAULT', 'PRIMARY', 'KEY',
+        'FOREIGN', 'REFERENCES', 'CHECK', 'UNIQUE', 'CONSTRAINT'
+    }
+    
+    @classmethod
+    def validate_identifier(cls, name: str, identifier_type: str = "identifier") -> str:
+        """
+        Validate a SQL identifier (database name, table name, column name, etc.)
+        
+        Args:
+            name: The identifier to validate
+            identifier_type: Type description for error messages (e.g., "database name", "table name")
+        
+        Returns:
+            The validated identifier (unchanged if valid)
+            
+        Raises:
+            SQLSecurityError: If the identifier is invalid
+        """
+        if not name:
+            raise SQLSecurityError(f"Empty {identifier_type} is not allowed")
+        
+        if not isinstance(name, str):
+            raise SQLSecurityError(f"Invalid {identifier_type}: must be a string, got {type(name).__name__}")
+        
+        # Strip whitespace
+        name = name.strip()
+        
+        if not name:
+            raise SQLSecurityError(f"Empty {identifier_type} is not allowed")
+        
+        # Check length
+        if len(name) > cls.MAX_IDENTIFIER_LENGTH:
+            raise SQLSecurityError(
+                f"Invalid {identifier_type}: '{name[:20]}...' exceeds maximum length of {cls.MAX_IDENTIFIER_LENGTH} characters"
+            )
+        
+        # Check for dangerous characters that could be SQL injection
+        dangerous_chars = ["'", '"', ';', '--', '/*', '*/', '\\', '\x00']
+        for char in dangerous_chars:
+            if char in name:
+                raise SQLSecurityError(
+                    f"Invalid {identifier_type}: '{name}' contains forbidden character '{char}'"
+                )
+        
+        # Validate pattern
+        if not cls.IDENTIFIER_PATTERN.match(name):
+            raise SQLSecurityError(
+                f"Invalid {identifier_type}: '{name}' contains invalid characters. "
+                f"Only letters, numbers, and underscores are allowed, and must start with a letter or underscore."
+            )
+        
+        logger.debug(f"Validated {identifier_type}: {name}")
+        return name
+    
+    @classmethod
+    def quote_identifier(cls, name: str, identifier_type: str = "identifier") -> str:
+        """
+        Safely quote a SQL identifier using backticks.
+        
+        Args:
+            name: The identifier to quote
+            identifier_type: Type description for error messages
+        
+        Returns:
+            The quoted identifier (e.g., `table_name`)
+            
+        Raises:
+            SQLSecurityError: If the identifier is invalid
+        """
+        # First validate the identifier
+        validated_name = cls.validate_identifier(name, identifier_type)
+        
+        # Escape any backticks within the name (double them)
+        escaped_name = validated_name.replace('`', '``')
+        
+        return f"`{escaped_name}`"
+    
+    @classmethod
+    def build_table_reference(
+        cls, 
+        table_name: str, 
+        db_name: Optional[str] = None, 
+        catalog_name: Optional[str] = None,
+        quote: bool = True
+    ) -> str:
+        """
+        Build a safe, fully-qualified table reference.
+        
+        Args:
+            table_name: The table name (required)
+            db_name: The database name (optional)
+            catalog_name: The catalog name (optional)
+            quote: Whether to quote identifiers with backticks (default: True)
+        
+        Returns:
+            A safe table reference string (e.g., `catalog`.`db`.`table`)
+            
+        Raises:
+            SQLSecurityError: If any identifier is invalid
+        """
+        parts = []
+        
+        if catalog_name:
+            if quote:
+                parts.append(cls.quote_identifier(catalog_name, "catalog name"))
+            else:
+                parts.append(cls.validate_identifier(catalog_name, "catalog name"))
+        
+        if db_name:
+            if quote:
+                parts.append(cls.quote_identifier(db_name, "database name"))
+            else:
+                parts.append(cls.validate_identifier(db_name, "database name"))
+        
+        if quote:
+            parts.append(cls.quote_identifier(table_name, "table name"))
+        else:
+            parts.append(cls.validate_identifier(table_name, "table name"))
+        
+        return '.'.join(parts)
+    
+    @classmethod
+    def build_column_reference(
+        cls,
+        column_name: str,
+        table_name: Optional[str] = None,
+        quote: bool = True
+    ) -> str:
+        """
+        Build a safe column reference.
+        
+        Args:
+            column_name: The column name (required)
+            table_name: The table name (optional, for qualified references)
+            quote: Whether to quote identifiers with backticks (default: True)
+        
+        Returns:
+            A safe column reference string (e.g., `table`.`column`)
+            
+        Raises:
+            SQLSecurityError: If any identifier is invalid
+        """
+        parts = []
+        
+        if table_name:
+            if quote:
+                parts.append(cls.quote_identifier(table_name, "table name"))
+            else:
+                parts.append(cls.validate_identifier(table_name, "table name"))
+        
+        if quote:
+            parts.append(cls.quote_identifier(column_name, "column name"))
+        else:
+            parts.append(cls.validate_identifier(column_name, "column name"))
+        
+        return '.'.join(parts)
+    
+    @classmethod
+    def validate_and_build_where_condition(
+        cls,
+        column_name: str,
+        operator: str = "=",
+        use_param: bool = True
+    ) -> Tuple[str, bool]:
+        """
+        Build a safe WHERE condition for a column.
+        
+        Args:
+            column_name: The column name
+            operator: The comparison operator (=, !=, <, >, <=, >=, LIKE, IN)
+            use_param: Whether to use parameterized placeholder (%s)
+        
+        Returns:
+            Tuple of (condition_string, needs_param)
+            e.g., ("`column` = %s", True) or ("`column` = DATABASE()", False)
+            
+        Raises:
+            SQLSecurityError: If column name is invalid or operator is not allowed
+        """
+        # Validate column name
+        quoted_column = cls.quote_identifier(column_name, "column name")
+        
+        # Validate operator
+        allowed_operators = {'=', '!=', '<>', '<', '>', '<=', '>=', 'LIKE', 'IN', 'IS'}
+        if operator.upper() not in allowed_operators:
+            raise SQLSecurityError(f"Invalid operator: '{operator}'. Allowed: {allowed_operators}")
+        
+        if use_param:
+            return f"{quoted_column} {operator} %s", True
+        else:
+            return f"{quoted_column} {operator}", False
+    
+    @staticmethod
+    def get_auth_context():
+        """
+        Get auth_context from the context variable.
+        
+        This retrieves the auth_context that was set by the HTTP middleware
+        during request processing.
+        
+        Returns:
+            The auth_context object, or None if not available
+        """
+        try:
+            auth_context = auth_context_var.get()
+            if auth_context:
+                logger.debug(f"Retrieved auth_context from context variable")
+            return auth_context
+        except Exception as e:
+            logger.debug(f"Could not retrieve auth_context: {e}")
+            return None
+    
+    @staticmethod
+    def set_auth_context(auth_context):
+        """
+        Set auth_context in the context variable.
+        
+        This is typically called by the HTTP middleware during request processing.
+        
+        Args:
+            auth_context: The auth_context object to set
+        """
+        auth_context_var.set(auth_context)
+        logger.debug("Set auth_context in context variable")
+
+
+# Convenience functions for direct use
+validate_identifier = SQLSecurityUtils.validate_identifier
+quote_identifier = SQLSecurityUtils.quote_identifier
+build_table_reference = SQLSecurityUtils.build_table_reference
+build_column_reference = SQLSecurityUtils.build_column_reference
+get_auth_context = SQLSecurityUtils.get_auth_context
+set_auth_context = SQLSecurityUtils.set_auth_context
+

pyproject.toml

@@ -20,7 +20,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "doris-mcp-server"
-version = "0.6.0"
+version = "0.6.1"
 description = "Enterprise-grade Model Context Protocol (MCP) server implementation for Apache Doris"
 authors = [
     {name = "Yijia Su", email = "freeoneplus@apache.org"}

start_server.sh

@@ -64,9 +64,10 @@ else
 fi
 
 # Set HTTP-specific environment variables
+# FIX for Issue #62 Bug 4: Use SERVER_PORT instead of MCP_PORT for consistency with code
 export MCP_TRANSPORT_TYPE="http"
 export MCP_HOST="${MCP_HOST:-0.0.0.0}"
-export MCP_PORT="${MCP_PORT:-3000}"
+export SERVER_PORT="${SERVER_PORT:-3000}"  # Changed from MCP_PORT to SERVER_PORT
 export WORKERS="${WORKERS:-1}"
 export ALLOWED_ORIGINS="${ALLOWED_ORIGINS:-*}"
 export LOG_LEVEL="${LOG_LEVEL:-info}"
@@ -77,15 +78,15 @@ export MCP_DEBUG_ADAPTER="true"
 export PYTHONPATH="$(pwd):$PYTHONPATH"
 
 echo -e "${GREEN}Starting MCP server (Streamable HTTP mode)...${NC}"
-echo -e "${YELLOW}Service will run on http://${MCP_HOST}:${MCP_PORT}/mcp${NC}"
-echo -e "${YELLOW}Health Check: http://${MCP_HOST}:${MCP_PORT}/health${NC}"
-echo -e "${YELLOW}MCP Endpoint: http://${MCP_HOST}:${MCP_PORT}/mcp${NC}"
-echo -e "${YELLOW}Local access: http://localhost:${MCP_PORT}/mcp${NC}"
+echo -e "${YELLOW}Service will run on http://${MCP_HOST}:${SERVER_PORT}/mcp${NC}"
+echo -e "${YELLOW}Health Check: http://${MCP_HOST}:${SERVER_PORT}/health${NC}"
+echo -e "${YELLOW}MCP Endpoint: http://${MCP_HOST}:${SERVER_PORT}/mcp${NC}"
+echo -e "${YELLOW}Local access: http://localhost:${SERVER_PORT}/mcp${NC}"
 echo -e "${YELLOW}Workers: ${WORKERS}${NC}"
 echo -e "${YELLOW}Use Ctrl+C to stop the service${NC}"
 
 # Start the server in HTTP mode (Streamable HTTP)
-python -m doris_mcp_server.main --transport http --host ${MCP_HOST} --port ${MCP_PORT} --workers ${WORKERS}
+python -m doris_mcp_server.main --transport http --host ${MCP_HOST} --port ${SERVER_PORT} --workers ${WORKERS}
 
 # Check exit status
 if [ $? -ne 0 ]; then
@@ -97,4 +98,4 @@ fi
 echo -e "${YELLOW}Tip: If the page displays abnormally, please clear your browser cache or use incognito mode${NC}"
 echo -e "${YELLOW}Chrome browser clear cache shortcut: Ctrl+Shift+Del (Windows) or Cmd+Shift+Del (Mac)${NC}"
 echo -e "${CYAN}For testing HTTP endpoints, you can use:${NC}"
-echo -e "${CYAN}  curl -X POST http://localhost:${MCP_PORT}/mcp -H 'Content-Type: application/json' -d '{\"method\":\"tools/list\"}'${NC}" 
+echo -e "${CYAN}  curl -X POST http://localhost:${SERVER_PORT}/mcp -H 'Content-Type: application/json' -d '{\"method\":\"tools/list\"}'${NC}" 

test/security/test_sql_injection.py

@@ -0,0 +1,367 @@
+#!/usr/bin/env python3
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+"""
+SQL Security Test Suite for Apache Doris MCP Server
+
+Tests for:
+1. SQL injection prevention via identifier validation
+2. Multi-statement SQL parsing in security validator
+3. auth_context enforcement
+"""
+
+import pytest
+import asyncio
+from unittest.mock import MagicMock, AsyncMock, patch
+
+
+class TestSQLSecurityUtils:
+    """Test cases for sql_security_utils module"""
+    
+    def test_validate_identifier_accepts_valid_names(self):
+        """Test that valid identifiers are accepted"""
+        from doris_mcp_server.utils.sql_security_utils import validate_identifier
+        
+        valid_names = [
+            "users",
+            "my_table",
+            "Table123",
+            "_private_table",
+            "CamelCaseTable",
+            "table_with_numbers_123",
+        ]
+        
+        for name in valid_names:
+            result = validate_identifier(name, "table")
+            assert result == name, f"Valid identifier '{name}' should be accepted"
+    
+    def test_validate_identifier_rejects_sql_injection(self):
+        """Test that SQL injection attempts are rejected"""
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier, 
+            SQLSecurityError
+        )
+        
+        injection_attempts = [
+            # Basic SQL injection
+            "'; DROP TABLE users; --",
+            "table' OR '1'='1",
+            "table'; DELETE FROM users; --",
+            
+            # Union-based injection
+            "table' UNION SELECT * FROM passwords --",
+            
+            # Comment injection
+            "table/**/OR/**/1=1",
+            "table--comment",
+            
+            # Special characters
+            "table`; DROP TABLE users;",
+            'table"; DROP TABLE users;',
+            "table\"; DELETE FROM",
+            
+            # Backtick escape attempt
+            "analytics`; SELECT * FROM sensitive_table;--",
+            
+            # Whitespace injection
+            "table name with spaces",
+            "table\ttab",
+            "table\nnewline",
+        ]
+        
+        for injection in injection_attempts:
+            with pytest.raises(SQLSecurityError):
+                validate_identifier(injection, "table")
+    
+    def test_validate_identifier_rejects_empty(self):
+        """Test that empty identifiers are rejected"""
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier, 
+            SQLSecurityError
+        )
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier("", "table")
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier(None, "table")
+    
+    def test_validate_identifier_rejects_too_long(self):
+        """Test that identifiers exceeding max length are rejected"""
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier, 
+            SQLSecurityError
+        )
+        
+        # Doris identifier max length is typically 64 characters
+        long_name = "a" * 100
+        with pytest.raises(SQLSecurityError):
+            validate_identifier(long_name, "table")
+    
+    def test_quote_identifier_adds_backticks(self):
+        """Test that quote_identifier properly escapes identifiers"""
+        from doris_mcp_server.utils.sql_security_utils import quote_identifier
+        
+        assert quote_identifier("my_table", "table") == "`my_table`"
+        assert quote_identifier("users", "table") == "`users`"
+        assert quote_identifier("Table123", "table") == "`Table123`"
+    
+    def test_quote_identifier_validates_first(self):
+        """Test that quote_identifier validates before quoting"""
+        from doris_mcp_server.utils.sql_security_utils import (
+            quote_identifier, 
+            SQLSecurityError
+        )
+        
+        with pytest.raises(SQLSecurityError):
+            quote_identifier("'; DROP TABLE users; --", "table")
+
+
+class TestSQLSecurityValidator:
+    """Test cases for SQLSecurityValidator multi-statement parsing"""
+    
+    @pytest.fixture
+    def dict_config(self):
+        """Create dictionary configuration"""
+        return {
+            "blocked_keywords": [
+                "DROP", "CREATE", "ALTER", "TRUNCATE",
+                "DELETE", "INSERT", "UPDATE",
+                "GRANT", "REVOKE", "EXEC", "EXECUTE"
+            ],
+            "max_query_complexity": 100,
+            "enable_security_check": True
+        }
+    
+    @pytest.fixture
+    def mock_auth_context(self):
+        """Create mock auth context"""
+        from doris_mcp_server.utils.security import AuthContext, SecurityLevel
+        return AuthContext(
+            user_id="test_user",
+            roles=["user"],
+            security_level=SecurityLevel.INTERNAL
+        )
+    
+    @pytest.mark.asyncio
+    async def test_validates_all_statements(self, dict_config, mock_auth_context):
+        """Test that validator checks ALL SQL statements, not just the first"""
+        from doris_mcp_server.utils.security import SQLSecurityValidator
+        
+        validator = SQLSecurityValidator(dict_config)
+        
+        # Multi-statement with injection in second statement
+        # This should be BLOCKED
+        malicious_sql = "SELECT 1; DROP TABLE users; SELECT 2"
+        
+        result = await validator.validate(malicious_sql, mock_auth_context)
+        
+        assert not result.is_valid, "Multi-statement injection should be blocked"
+        # Check for either DROP keyword detection or SQL injection detection
+        error_upper = result.error_message.upper()
+        assert ("DROP" in error_upper or 
+                "INJECTION" in error_upper or 
+                "BLOCKED" in error_upper), f"Expected DROP/injection/blocked in: {result.error_message}"
+    
+    @pytest.mark.asyncio
+    async def test_blocks_hidden_dangerous_statement(self, dict_config, mock_auth_context):
+        """Test that dangerous statements hidden after safe ones are blocked"""
+        from doris_mcp_server.utils.security import SQLSecurityValidator
+        
+        validator = SQLSecurityValidator(dict_config)
+        
+        # Safe statement followed by dangerous one
+        malicious_sql = """
+        SELECT * FROM users WHERE id = 1;
+        DELETE FROM audit_log;
+        SELECT 1;
+        """
+        
+        result = await validator.validate(malicious_sql, mock_auth_context)
+        
+        assert not result.is_valid, "Hidden DELETE statement should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_allows_safe_multi_statement(self, dict_config, mock_auth_context):
+        """Test that multiple safe SELECT statements are allowed"""
+        from doris_mcp_server.utils.security import SQLSecurityValidator
+        
+        validator = SQLSecurityValidator(dict_config)
+        
+        safe_sql = """
+        SELECT * FROM users;
+        SELECT COUNT(*) FROM orders;
+        SELECT id, name FROM products;
+        """
+        
+        result = await validator.validate(safe_sql, mock_auth_context)
+        
+        assert result.is_valid, f"Multiple safe SELECT statements should be allowed, got: {result.error_message}"
+    
+    @pytest.mark.asyncio
+    async def test_context_switch_injection_blocked(self, dict_config, mock_auth_context):
+        """Test that context switch SQL injection is blocked"""
+        from doris_mcp_server.utils.security import SQLSecurityValidator
+        
+        validator = SQLSecurityValidator(dict_config)
+        
+        # Simulating the exec_query_for_mcp attack vector
+        injected_sql = """
+        USE `analytics`; SELECT * FROM sensitive_table;-- `;
+        SELECT * FROM public_table;
+        """
+        
+        result = await validator.validate(injected_sql, mock_auth_context)
+        
+        # The validator should process all statements
+        # Even if USE is allowed, subsequent unauthorized access should be caught
+        # by table access checks (if configured)
+
+
+class TestExecQueryForMCP:
+    """Test cases for exec_query_for_mcp function"""
+    
+    @pytest.mark.asyncio
+    async def test_rejects_malicious_db_name(self):
+        """Test that malicious db_name is rejected"""
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier,
+            SQLSecurityError
+        )
+        
+        # The attack vector from security report
+        malicious_db_name = "analytics`; SELECT * FROM sensitive_table;--"
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier(malicious_db_name, "database name")
+    
+    @pytest.mark.asyncio
+    async def test_rejects_malicious_catalog_name(self):
+        """Test that malicious catalog_name is rejected"""
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier,
+            SQLSecurityError
+        )
+        
+        malicious_catalog_name = "internal'; DROP DATABASE production;--"
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier(malicious_catalog_name, "catalog name")
+
+
+class TestDependencyAnalysisTools:
+    """Test cases for dependency_analysis_tools security fixes"""
+    
+    @pytest.mark.asyncio
+    async def test_get_tables_metadata_rejects_injection(self):
+        """Test that _get_tables_metadata rejects SQL injection"""
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier,
+            SQLSecurityError
+        )
+        
+        # The attack vector from security report
+        injection_db_name = "test_db' OR '1'='1' --"
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier(injection_db_name, "database name")
+
+
+class TestAuthContextEnforcement:
+    """Test cases for auth_context enforcement"""
+    
+    def test_execute_requires_auth_context_for_security(self):
+        """Test that security checks require auth_context"""
+        # This test documents the expected behavior:
+        # When auth_context is None, security checks are skipped
+        # When auth_context is provided, security checks are performed
+        
+        # The fix ensures all execute() calls pass auth_context
+        pass
+    
+    @pytest.mark.asyncio
+    async def test_get_auth_context_returns_context(self):
+        """Test that get_auth_context retrieves context from ContextVar"""
+        from doris_mcp_server.utils.sql_security_utils import get_auth_context
+        
+        # When no context is set, should return None
+        result = get_auth_context()
+        # This is expected - context is set by HTTP middleware
+        assert result is None or hasattr(result, 'user_id')
+
+
+class TestIntegrationScenarios:
+    """Integration test scenarios for security fixes"""
+    
+    def test_attack_scenario_1_permission_bypass(self):
+        """
+        Attack Scenario 1: Permission Bypass in Multi-Tenant Environment
+        
+        Expected: User can only query their own database (db_name="tenant_a_db")
+        Attack: Inject "tenant_a_db' OR '1'='1' --" to query ALL databases
+        Result: Should be BLOCKED by validate_identifier()
+        """
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier,
+            SQLSecurityError
+        )
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier("tenant_a_db' OR '1'='1' --", "database name")
+    
+    def test_attack_scenario_2_union_injection(self):
+        """
+        Attack Scenario 2: UNION-based Information Disclosure
+        
+        Attack: Inject UNION SELECT to extract sensitive data
+        Result: Should be BLOCKED
+        """
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier,
+            SQLSecurityError
+        )
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier(
+                "test' UNION SELECT password FROM users --",
+                "database name"
+            )
+    
+    def test_attack_scenario_3_backtick_escape(self):
+        """
+        Attack Scenario 3: Backtick Escape Attempt
+        
+        Attack: Use backticks to break out of quoted identifier
+        Result: Should be BLOCKED
+        """
+        from doris_mcp_server.utils.sql_security_utils import (
+            validate_identifier,
+            SQLSecurityError
+        )
+        
+        with pytest.raises(SQLSecurityError):
+            validate_identifier(
+                "analytics`; SELECT * FROM sensitive_table;--",
+                "database name"
+            )
+
+
+# Run tests with: pytest tests/test_sql_security.py -v
+if __name__ == "__main__":
+    pytest.main([__file__, "-v", "--tb=short"])
+

test/security/test_sql_injection_api.py

@@ -0,0 +1,871 @@
+#!/usr/bin/env python3
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+"""
+SQL Injection API Integration Tests
+
+This module tests SQL injection prevention through the MCP HTTP API.
+It sends malicious payloads and verifies they are properly blocked.
+
+Prerequisites:
+    - MCP server running on localhost:3000
+    - Run with: pytest test/security/test_sql_injection_api.py -v
+
+Usage:
+    # Start server first
+    bash start_server.sh
+    
+    # Run tests
+    pytest test/security/test_sql_injection_api.py -v --no-cov
+"""
+
+import pytest
+import httpx
+import json
+import asyncio
+from typing import Optional
+
+
+# Server configuration
+MCP_BASE_URL = "http://localhost:3000"
+MCP_ENDPOINT = f"{MCP_BASE_URL}/mcp"
+HEALTH_ENDPOINT = f"{MCP_BASE_URL}/health"
+TIMEOUT = 30.0
+
+
+class MCPClient:
+    """Simple MCP HTTP client for testing"""
+    
+    def __init__(self, base_url: str = MCP_BASE_URL):
+        self.base_url = base_url
+        self.mcp_endpoint = f"{base_url}/mcp"
+        self.session_id: Optional[str] = None
+        self.request_id = 0
+        self.client = httpx.AsyncClient(timeout=TIMEOUT)
+    
+    async def close(self):
+        await self.client.aclose()
+    
+    def _next_id(self) -> int:
+        self.request_id += 1
+        return self.request_id
+    
+    async def initialize(self) -> dict:
+        """Initialize MCP session"""
+        response = await self.client.post(
+            self.mcp_endpoint,
+            headers={
+                "Content-Type": "application/json",
+                "Accept": "application/json, text/event-stream"
+            },
+            json={
+                "jsonrpc": "2.0",
+                "method": "initialize",
+                "params": {
+                    "protocolVersion": "2024-11-05",
+                    "capabilities": {},
+                    "clientInfo": {
+                        "name": "sql-injection-test",
+                        "version": "1.0.0"
+                    }
+                },
+                "id": self._next_id()
+            }
+        )
+        
+        # Extract session ID from response header
+        self.session_id = response.headers.get("mcp-session-id")
+        return self._parse_response(response.text)
+    
+    async def call_tool(self, tool_name: str, arguments: dict) -> dict:
+        """Call an MCP tool"""
+        if not self.session_id:
+            await self.initialize()
+        
+        response = await self.client.post(
+            self.mcp_endpoint,
+            headers={
+                "Content-Type": "application/json",
+                "Accept": "application/json, text/event-stream",
+                "mcp-session-id": self.session_id
+            },
+            json={
+                "jsonrpc": "2.0",
+                "method": "tools/call",
+                "params": {
+                    "name": tool_name,
+                    "arguments": arguments
+                },
+                "id": self._next_id()
+            }
+        )
+        
+        return self._parse_response(response.text)
+    
+    def _parse_response(self, text: str) -> dict:
+        """Parse JSON response"""
+        try:
+            return json.loads(text)
+        except json.JSONDecodeError:
+            # Try SSE format
+            lines = text.strip().split("\n")
+            for line in lines:
+                if line.startswith("data: "):
+                    try:
+                        return json.loads(line[6:])
+                    except json.JSONDecodeError:
+                        continue
+            return {"raw": text}
+
+
+def print_result(test_name: str, payload: dict, result: dict):
+    """Print test result in a readable format"""
+    print(f"\n{'='*60}")
+    print(f"TEST: {test_name}")
+    print(f"{'='*60}")
+    print(f"PAYLOAD: {json.dumps(payload, ensure_ascii=False)}")
+    print(f"{'-'*60}")
+    
+    # Extract inner result content
+    if "result" in result and "content" in result.get("result", {}):
+        for item in result["result"]["content"]:
+            if item.get("type") == "text":
+                try:
+                    inner = json.loads(item["text"])
+                    print("RESPONSE:")
+                    print(f"  success: {inner.get('success')}")
+                    if inner.get('error'):
+                        print(f"  error: {inner.get('error')}")
+                    if inner.get('error_type'):
+                        print(f"  error_type: {inner.get('error_type')}")
+                    if inner.get('risk_level'):
+                        print(f"  risk_level: {inner.get('risk_level')}")
+                    if inner.get('message'):
+                        print(f"  message: {inner.get('message')}")
+                    if inner.get('data') is not None and inner.get('success'):
+                        data_str = json.dumps(inner.get('data'), ensure_ascii=False)
+                        if len(data_str) > 200:
+                            data_str = data_str[:200] + "..."
+                        print(f"  data: {data_str}")
+                except (json.JSONDecodeError, TypeError):
+                    print(f"RESPONSE (raw): {item.get('text', '')[:500]}")
+    elif "error" in result:
+        print(f"RESPONSE ERROR: {result['error']}")
+    else:
+        print(f"RESPONSE (raw): {json.dumps(result, ensure_ascii=False)[:500]}")
+    
+    print(f"{'='*60}\n")
+
+
+class TestSQLInjectionAPI:
+    """Test SQL injection prevention through MCP API"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.fixture
+    def is_server_running(self):
+        """Check if MCP server is running"""
+        import httpx
+        try:
+            response = httpx.get(HEALTH_ENDPOINT, timeout=5.0)
+            return response.status_code == 200
+        except Exception:
+            return False
+    
+    @pytest.mark.asyncio
+    async def test_server_health(self):
+        """Test that MCP server is running and healthy"""
+        async with httpx.AsyncClient() as client:
+            response = await client.get(HEALTH_ENDPOINT)
+            assert response.status_code == 200
+            data = response.json()
+            assert data["status"] == "healthy"
+    
+    @pytest.mark.asyncio
+    async def test_exec_query_with_drop_injection(self, mcp_client):
+        """Test exec_query rejects DROP TABLE injection"""
+        # Classic SQL injection: append DROP TABLE
+        payload = {"sql": "SELECT * FROM users; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("DROP TABLE Injection", payload, result)
+        
+        # Should return error, not execute the DROP
+        assert self._is_blocked_or_error(result), \
+            f"DROP TABLE injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_exec_query_with_union_injection(self, mcp_client):
+        """Test exec_query blocks UNION-based injection attempts"""
+        # UNION injection to extract data from other tables
+        payload = {"sql": "SELECT id FROM users UNION SELECT password FROM admin_users"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("UNION Injection", payload, result)
+    
+    @pytest.mark.asyncio
+    async def test_exec_query_with_delete_injection(self, mcp_client):
+        """Test exec_query rejects DELETE injection"""
+        payload = {"sql": "SELECT 1; DELETE FROM users WHERE 1=1; SELECT 2"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("DELETE Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"DELETE injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_exec_query_with_update_injection(self, mcp_client):
+        """Test exec_query rejects UPDATE injection"""
+        payload = {"sql": "SELECT 1; UPDATE users SET role='admin' WHERE id=1; --"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("UPDATE Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"UPDATE injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_exec_query_db_name_injection(self, mcp_client):
+        """Test exec_query rejects SQL injection via db_name parameter"""
+        # Attack vector: inject SQL via db_name parameter
+        payload = {"sql": "SELECT 1", "db_name": "test'; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("db_name Parameter Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"db_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_exec_query_catalog_name_injection(self, mcp_client):
+        """Test exec_query rejects SQL injection via catalog_name parameter"""
+        # Attack vector: inject SQL via catalog_name parameter
+        payload = {"sql": "SELECT 1", "catalog_name": "internal`; SELECT * FROM mysql.user; --"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("catalog_name Parameter Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"catalog_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_table_schema_injection(self, mcp_client):
+        """Test get_table_schema rejects SQL injection via table_name"""
+        # Attack vector: inject SQL via table_name parameter
+        payload = {"table_name": "users'; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("get_table_schema", payload)
+        print_result("table_name Injection (get_table_schema)", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_table_schema_db_injection(self, mcp_client):
+        """Test get_table_schema rejects SQL injection via db_name"""
+        payload = {"table_name": "users", "db_name": "test' OR '1'='1"}
+        result = await mcp_client.call_tool("get_table_schema", payload)
+        print_result("db_name Injection (get_table_schema)", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"db_name injection in get_table_schema should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_analyze_dependencies_injection(self, mcp_client):
+        """Test analyze_dependencies rejects SQL injection"""
+        # This was the original vulnerability reported
+        payload = {"table_name": "users", "db_name": "test_db' OR '1'='1' --"}
+        result = await mcp_client.call_tool("analyze_dependencies", payload)
+        print_result("analyze_dependencies Injection (Original Report)", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"analyze_dependencies db_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_stacked_queries_injection(self, mcp_client):
+        """Test that stacked queries (multiple statements) are blocked"""
+        # Multiple statements injection
+        payload = {"sql": "SELECT * FROM users WHERE id = 1; INSERT INTO audit_log VALUES (NULL, 'hacked', NOW()); SELECT 1;"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Stacked Queries (INSERT) Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"Stacked queries with INSERT should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_comment_based_injection(self, mcp_client):
+        """Test that comment-based injection is blocked"""
+        # Using comments to bypass filters
+        payload = {"sql": "SELECT * FROM users WHERE id = 1/**/OR/**/1=1"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Comment-based Injection", payload, result)
+    
+    @pytest.mark.asyncio
+    async def test_hex_encoded_injection(self, mcp_client):
+        """Test that hex-encoded injection attempts are handled"""
+        # Hex-encoded 'DROP' attempt
+        payload = {"sql": "SELECT 0x44524F50205441424C4520757365727320"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Hex Encoded Injection", payload, result)
+    
+    @pytest.mark.asyncio
+    async def test_backtick_escape_injection(self, mcp_client):
+        """Test backtick escape injection is blocked"""
+        # Attempt to escape backtick quoting
+        payload = {"sql": "SELECT 1", "db_name": "analytics`; SELECT * FROM sensitive_table;--"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Backtick Escape Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            f"Backtick escape injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_valid_query_succeeds(self, mcp_client):
+        """Test that valid queries still work"""
+        # Simple valid query should work
+        payload = {"sql": "SELECT 1 AS test_value"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Valid Query (should succeed)", payload, result)
+    
+    @pytest.mark.asyncio
+    async def test_valid_show_databases(self, mcp_client):
+        """Test that SHOW DATABASES works"""
+        payload = {"sql": "SHOW DATABASES"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("SHOW DATABASES (should succeed)", payload, result)
+    
+    def _is_blocked_or_error(self, result: dict) -> bool:
+        """Check if result indicates blocked or error"""
+        if not result:
+            return True
+        
+        # Check for JSON-RPC error
+        if "error" in result:
+            return True
+        
+        # Check for error in result content
+        if "result" in result:
+            result_content = result.get("result", {})
+            if isinstance(result_content, dict):
+                # Check for isError flag
+                if result_content.get("isError"):
+                    return True
+                # Check content array for error messages
+                content = result_content.get("content", [])
+                for item in content:
+                    if isinstance(item, dict):
+                        text = item.get("text", "")
+                        # Parse the JSON text content
+                        try:
+                            text_data = json.loads(text)
+                            # Check for success: false
+                            if text_data.get("success") is False:
+                                return True
+                            # Check for error field
+                            if text_data.get("error"):
+                                return True
+                        except (json.JSONDecodeError, TypeError):
+                            pass
+                        # Check text for security keywords
+                        if any(keyword in text.lower() for keyword in [
+                            "error", "blocked", "invalid", "security",
+                            "injection", "denied", "forbidden", "not allowed",
+                            "security_violation", "risk_level"
+                        ]):
+                            return True
+        
+        # Check raw text response
+        raw = result.get("raw", "")
+        if isinstance(raw, str) and any(keyword in raw.lower() for keyword in [
+            "error", "blocked", "invalid", "security"
+        ]):
+            return True
+        
+        return False
+
+
+class TestIdentifierInjectionAPI:
+    """Test identifier-based SQL injection prevention"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.mark.asyncio
+    async def test_table_name_with_semicolon(self, mcp_client):
+        """Test table name containing semicolon is rejected"""
+        payload = {"table_name": "users; DROP TABLE users"}
+        result = await mcp_client.call_tool("get_table_schema", payload)
+        print_result("Table Name with Semicolon", payload, result)
+        
+        # Should be blocked by identifier validation
+        assert self._contains_error_indicator(result), \
+            f"Table name with semicolon should be rejected"
+    
+    @pytest.mark.asyncio
+    async def test_table_name_with_quotes(self, mcp_client):
+        """Test table name containing quotes is rejected"""
+        payload = {"table_name": "users' OR '1'='1"}
+        result = await mcp_client.call_tool("get_table_schema", payload)
+        print_result("Table Name with Quotes", payload, result)
+        
+        assert self._contains_error_indicator(result), \
+            f"Table name with quotes should be rejected"
+    
+    @pytest.mark.asyncio  
+    async def test_db_name_with_special_chars(self, mcp_client):
+        """Test database name with special characters is rejected"""
+        special_chars = [
+            "test;db",
+            "test'db",
+            "test\"db",
+            "test`db",
+            "test--db",
+            "test/*db*/",
+        ]
+        
+        for db_name in special_chars:
+            payload = {"table_name": "users", "db_name": db_name}
+            result = await mcp_client.call_tool("get_table_schema", payload)
+            print_result(f"Special Char in db_name: {db_name}", payload, result)
+            
+            assert self._contains_error_indicator(result), \
+                f"db_name '{db_name}' should be rejected"
+    
+    @pytest.mark.asyncio
+    async def test_valid_identifiers_accepted(self, mcp_client):
+        """Test that valid identifiers are accepted"""
+        valid_names = [
+            "users",
+            "my_table",
+            "Table123",
+            "_internal_table",
+        ]
+        
+        for table_name in valid_names:
+            payload = {"table_name": table_name}
+            result = await mcp_client.call_tool("get_table_schema", payload)
+            print_result(f"Valid Identifier: {table_name}", payload, result)
+    
+    def _contains_error_indicator(self, result: dict) -> bool:
+        """Check if result contains error indicators"""
+        if not result:
+            return True
+        
+        # Check for JSON-RPC error
+        if "error" in result:
+            return True
+        
+        # Check result content
+        result_str = json.dumps(result).lower()
+        error_keywords = [
+            "error", "invalid", "illegal", "blocked",
+            "security", "injection", "denied", "forbidden"
+        ]
+        
+        return any(keyword in result_str for keyword in error_keywords)
+
+
+class TestMultiStatementInjectionAPI:
+    """Test multi-statement SQL injection prevention"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.mark.asyncio
+    async def test_hidden_drop_after_select(self, mcp_client):
+        """Test DROP hidden after legitimate SELECT is blocked"""
+        payload = {"sql": "SELECT id, name FROM users WHERE status = 'active'; DROP TABLE audit_log; SELECT 1;"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Hidden DROP after SELECT", payload, result)
+        
+        assert self._is_dangerous_blocked(result), \
+            f"Hidden DROP statement should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_hidden_truncate_after_select(self, mcp_client):
+        """Test TRUNCATE hidden after SELECT is blocked"""
+        payload = {"sql": "SELECT 1; TRUNCATE TABLE users"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Hidden TRUNCATE after SELECT", payload, result)
+        
+        assert self._is_dangerous_blocked(result), \
+            f"Hidden TRUNCATE should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_hidden_grant_after_select(self, mcp_client):
+        """Test GRANT hidden after SELECT is blocked"""
+        payload = {"sql": "SELECT 1; GRANT ALL ON *.* TO 'hacker'@'%'"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Hidden GRANT after SELECT", payload, result)
+        
+        assert self._is_dangerous_blocked(result), \
+            f"Hidden GRANT should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_multiple_safe_selects_allowed(self, mcp_client):
+        """Test that multiple SELECT statements may be allowed"""
+        payload = {"sql": "SELECT 1; SELECT 2; SELECT 3;"}
+        result = await mcp_client.call_tool("exec_query", payload)
+        print_result("Multiple Safe SELECTs", payload, result)
+    
+    def _is_dangerous_blocked(self, result: dict) -> bool:
+        """Check if dangerous operation was blocked"""
+        if not result:
+            return True
+        
+        # Check for error
+        if "error" in result:
+            return True
+        
+        # Check result content for blocking indicators
+        result_str = json.dumps(result).lower()
+        block_indicators = [
+            "drop", "truncate", "grant", "revoke",
+            "blocked", "denied", "forbidden", "not allowed",
+            "security", "error"
+        ]
+        
+        return any(indicator in result_str for indicator in block_indicators)
+
+
+class TestADBCQueryInjectionAPI:
+    """Test ADBC query SQL injection prevention"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.mark.asyncio
+    async def test_exec_adbc_query_drop_injection(self, mcp_client):
+        """Test exec_adbc_query rejects DROP TABLE injection"""
+        payload = {"sql": "SELECT * FROM users; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("exec_adbc_query", payload)
+        print_result("ADBC DROP TABLE Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "ADBC DROP TABLE injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_exec_adbc_query_delete_injection(self, mcp_client):
+        """Test exec_adbc_query rejects DELETE injection"""
+        payload = {"sql": "SELECT 1; DELETE FROM users; --"}
+        result = await mcp_client.call_tool("exec_adbc_query", payload)
+        print_result("ADBC DELETE Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "ADBC DELETE injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_exec_adbc_query_valid(self, mcp_client):
+        """Test exec_adbc_query allows valid queries"""
+        payload = {"sql": "SELECT 1 AS test"}
+        result = await mcp_client.call_tool("exec_adbc_query", payload)
+        print_result("ADBC Valid Query", payload, result)
+    
+    def _is_blocked_or_error(self, result: dict) -> bool:
+        """Check if result indicates blocked or error"""
+        if not result:
+            return True
+        if "error" in result:
+            return True
+        result_str = json.dumps(result).lower()
+        return any(kw in result_str for kw in ["error", "blocked", "invalid", "security", "injection"])
+
+
+class TestMetadataToolsInjectionAPI:
+    """Test metadata tools SQL injection prevention"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.mark.asyncio
+    async def test_get_db_table_list_db_injection(self, mcp_client):
+        """Test get_db_table_list rejects db_name injection"""
+        payload = {"db_name": "test'; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("get_db_table_list", payload)
+        print_result("get_db_table_list db_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "db_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_db_table_list_catalog_injection(self, mcp_client):
+        """Test get_db_table_list rejects catalog_name injection"""
+        payload = {"catalog_name": "internal`; SELECT * FROM mysql.user; --"}
+        result = await mcp_client.call_tool("get_db_table_list", payload)
+        print_result("get_db_table_list catalog_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "catalog_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_table_comment_injection(self, mcp_client):
+        """Test get_table_comment rejects table_name injection"""
+        payload = {"table_name": "users'; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("get_table_comment", payload)
+        print_result("get_table_comment table_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_table_column_comments_injection(self, mcp_client):
+        """Test get_table_column_comments rejects injection"""
+        payload = {"table_name": "users'; DROP TABLE users; --", "db_name": "test"}
+        result = await mcp_client.call_tool("get_table_column_comments", payload)
+        print_result("get_table_column_comments Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_table_indexes_injection(self, mcp_client):
+        """Test get_table_indexes rejects table_name injection"""
+        payload = {"table_name": "users; DROP TABLE users", "db_name": "test"}
+        result = await mcp_client.call_tool("get_table_indexes", payload)
+        print_result("get_table_indexes Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    def _is_blocked_or_error(self, result: dict) -> bool:
+        """Check if result indicates blocked or error"""
+        if not result:
+            return True
+        if "error" in result:
+            return True
+        result_str = json.dumps(result).lower()
+        return any(kw in result_str for kw in ["error", "blocked", "invalid", "security", "injection"])
+
+
+class TestAnalyticsToolsInjectionAPI:
+    """Test analytics tools SQL injection prevention"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.mark.asyncio
+    async def test_analyze_columns_table_injection(self, mcp_client):
+        """Test analyze_columns rejects table_name injection"""
+        payload = {"table_name": "users'; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("analyze_columns", payload)
+        print_result("analyze_columns table_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_analyze_columns_db_injection(self, mcp_client):
+        """Test analyze_columns rejects db_name injection"""
+        payload = {"table_name": "users", "db_name": "test' OR '1'='1"}
+        result = await mcp_client.call_tool("analyze_columns", payload)
+        print_result("analyze_columns db_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "db_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_table_basic_info_injection(self, mcp_client):
+        """Test get_table_basic_info rejects injection"""
+        payload = {"table_name": "users; DROP TABLE audit_log"}
+        result = await mcp_client.call_tool("get_table_basic_info", payload)
+        print_result("get_table_basic_info Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_analyze_table_storage_injection(self, mcp_client):
+        """Test analyze_table_storage rejects injection"""
+        payload = {"table_name": "users`; SELECT * FROM sensitive; --"}
+        result = await mcp_client.call_tool("analyze_table_storage", payload)
+        print_result("analyze_table_storage Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_sql_explain_injection(self, mcp_client):
+        """Test get_sql_explain rejects SQL injection"""
+        payload = {"sql": "SELECT 1; DROP TABLE users; --"}
+        result = await mcp_client.call_tool("get_sql_explain", payload)
+        print_result("get_sql_explain SQL Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "SQL injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_sql_profile_injection(self, mcp_client):
+        """Test get_sql_profile rejects SQL injection"""
+        payload = {"sql": "SELECT 1; DELETE FROM audit_log; --"}
+        result = await mcp_client.call_tool("get_sql_profile", payload)
+        print_result("get_sql_profile SQL Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "SQL injection should be blocked"
+    
+    def _is_blocked_or_error(self, result: dict) -> bool:
+        """Check if result indicates blocked or error"""
+        if not result:
+            return True
+        if "error" in result:
+            return True
+        result_str = json.dumps(result).lower()
+        return any(kw in result_str for kw in ["error", "blocked", "invalid", "security", "injection"])
+
+
+class TestGovernanceToolsInjectionAPI:
+    """Test data governance tools SQL injection prevention"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.mark.asyncio
+    async def test_trace_column_lineage_table_injection(self, mcp_client):
+        """Test trace_column_lineage rejects table_name injection"""
+        payload = {"table_name": "users'; DROP TABLE users; --", "column_name": "id"}
+        result = await mcp_client.call_tool("trace_column_lineage", payload)
+        print_result("trace_column_lineage table_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_trace_column_lineage_column_injection(self, mcp_client):
+        """Test trace_column_lineage rejects column_name injection"""
+        payload = {"table_name": "users", "column_name": "id; DROP TABLE users"}
+        result = await mcp_client.call_tool("trace_column_lineage", payload)
+        print_result("trace_column_lineage column_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "column_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_monitor_data_freshness_injection(self, mcp_client):
+        """Test monitor_data_freshness rejects table_name injection"""
+        payload = {"table_name": "users`; SELECT * FROM passwords; --"}
+        result = await mcp_client.call_tool("monitor_data_freshness", payload)
+        print_result("monitor_data_freshness Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_analyze_data_access_patterns_injection(self, mcp_client):
+        """Test analyze_data_access_patterns rejects injection"""
+        payload = {"table_name": "users' UNION SELECT password FROM admin --"}
+        result = await mcp_client.call_tool("analyze_data_access_patterns", payload)
+        print_result("analyze_data_access_patterns Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    def _is_blocked_or_error(self, result: dict) -> bool:
+        """Check if result indicates blocked or error"""
+        if not result:
+            return True
+        if "error" in result:
+            return True
+        result_str = json.dumps(result).lower()
+        return any(kw in result_str for kw in ["error", "blocked", "invalid", "security", "injection"])
+
+
+class TestPerformanceToolsInjectionAPI:
+    """Test performance analytics tools SQL injection prevention"""
+    
+    @pytest.fixture
+    async def mcp_client(self):
+        """Create MCP client instance"""
+        client = MCPClient()
+        yield client
+        await client.close()
+    
+    @pytest.mark.asyncio
+    async def test_analyze_slow_queries_db_injection(self, mcp_client):
+        """Test analyze_slow_queries_topn rejects db_name injection"""
+        payload = {"db_name": "test'; DROP TABLE audit_log; --"}
+        result = await mcp_client.call_tool("analyze_slow_queries_topn", payload)
+        print_result("analyze_slow_queries_topn db_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "db_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_analyze_resource_growth_db_injection(self, mcp_client):
+        """Test analyze_resource_growth_curves rejects db_name injection"""
+        payload = {"db_name": "test`; GRANT ALL ON *.* TO 'hacker'; --"}
+        result = await mcp_client.call_tool("analyze_resource_growth_curves", payload)
+        print_result("analyze_resource_growth_curves db_name Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "db_name injection should be blocked"
+    
+    @pytest.mark.asyncio
+    async def test_get_table_data_size_injection(self, mcp_client):
+        """Test get_table_data_size rejects table_name injection"""
+        payload = {"table_name": "users; TRUNCATE TABLE logs"}
+        result = await mcp_client.call_tool("get_table_data_size", payload)
+        print_result("get_table_data_size Injection", payload, result)
+        
+        assert self._is_blocked_or_error(result), \
+            "table_name injection should be blocked"
+    
+    def _is_blocked_or_error(self, result: dict) -> bool:
+        """Check if result indicates blocked or error"""
+        if not result:
+            return True
+        if "error" in result:
+            return True
+        result_str = json.dumps(result).lower()
+        return any(kw in result_str for kw in ["error", "blocked", "invalid", "security", "injection"])
+
+
+# Pytest configuration for async tests
+@pytest.fixture(scope="session")
+def event_loop():
+    """Create event loop for async tests"""
+    loop = asyncio.new_event_loop()
+    yield loop
+    loop.close()
+
+
+if __name__ == "__main__":
+    pytest.main([__file__, "-v", "--tb=short", "-x"])
+

test/utils/test_query_executor.py

@@ -201,3 +201,73 @@ class TestDorisQueryExecutor:
             if result["success"]:
                 assert "data" in result
                 assert "row_count" in result 
+
+    @pytest.mark.asyncio
+    async def test_execute_multi_sql_statements(self, query_executor):
+        """Test execution of multiple SQL statements"""
+        from doris_mcp_server.utils.query_executor import QueryResult
+        
+        # Disable security check for this test
+        query_executor.connection_manager.config.security.enable_security_check = False
+        
+        with patch.object(query_executor, 'execute_query') as mock_execute:
+            # Mock results for three SQL statements
+            mock_execute.side_effect = [
+                QueryResult(
+                    data=[{"id": 1, "name": "张三"}],
+                    row_count=1,
+                    execution_time=0.1,
+                    sql="SELECT id, name FROM users WHERE id = 1",
+                    metadata={"columns": ["id", "name"]}
+                ),
+                QueryResult(
+                    data=[{"id": 2, "name": "李四"}],
+                    row_count=1,
+                    execution_time=0.12,
+                    sql="SELECT id, name FROM users WHERE id = 2",
+                    metadata={"columns": ["id", "name"]}
+                ),
+                QueryResult(
+                    data=[{"count": 100}],
+                    row_count=1,
+                    execution_time=0.08,
+                    sql="SELECT COUNT(*) as count FROM users",
+                    metadata={"columns": ["count"]}
+                )
+            ]
+            
+            # Execute multiple SQL statements separated by semicolons
+            multi_sql = """
+                SELECT id, name FROM users WHERE id = 1;
+                SELECT id, name FROM users WHERE id = 2;
+                SELECT COUNT(*) as count FROM users;
+            """
+            
+            result = await query_executor.execute_sql_for_mcp(multi_sql)
+            
+            # Verify the result structure for multiple statements
+            assert result["success"] is True
+            assert result["multiple_results"] is True
+            assert "results" in result
+            assert len(result["results"]) == 3
+            
+            # Verify first query result
+            assert result["results"][0]["data"] == [{"id": 1, "name": "张三"}]
+            assert result["results"][0]["row_count"] == 1
+            assert result["results"][0]["metadata"]["columns"] == ["id", "name"]
+            assert result["results"][0]["metadata"]["query"] == "SELECT id, name FROM users WHERE id = 1"
+            
+            # Verify second query result
+            assert result["results"][1]["data"] == [{"id": 2, "name": "李四"}]
+            assert result["results"][1]["row_count"] == 1
+            assert result["results"][1]["metadata"]["columns"] == ["id", "name"]
+            assert result["results"][1]["metadata"]["query"] == "SELECT id, name FROM users WHERE id = 2"
+            
+            # Verify third query result
+            assert result["results"][2]["data"] == [{"count": 100}]
+            assert result["results"][2]["row_count"] == 1
+            assert result["results"][2]["metadata"]["columns"] == ["count"]
+            assert result["results"][2]["metadata"]["query"] == "SELECT COUNT(*) as count FROM users"
+            
+            # Verify execute_query was called three times
+            assert mock_execute.call_count == 3

uv.lock

@@ -562,7 +562,7 @@ wheels = [
 
 [[package]]
 name = "doris-mcp-server"
-version = "0.5.1"
+version = "0.6.1"
 source = { editable = "." }
 dependencies = [
     { name = "adbc-driver-flightsql" },