fix: resolve MariaDB collation error by switching from mysql-connector to PyMySQL

- Replace mysql-connector-python with PyMySQL driver for better MariaDB compatibility
- PyMySQL handles utf8mb4_0900_ai_ci collation properly without errors
- Update Dockerfile.sql-executor to install PyMySQL and psycopg2-binary
- Refactor sql-query-executor.py to use PyMySQL API (pymysql.connect, DictCursor)
- Verified sql-executor service with SELECT, INSERT, UPDATE operations on Freescout DB
- Add n8n workflow definitions: workflow-a-http.json and workflow-b-http.json
  * Workflow A: Polls unprocessed conversations, analyzes with LiteLLM, saves suggestions
  * Workflow B: Polls approved suggestions, executes Baramundi jobs or email replies
- Update compose.yaml with sql-executor service configuration and dependencies

All SQL operations now execute successfully against MariaDB 11.3.2

Dockerfile.sql-executor

@@ -0,0 +1,19 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+# Install dependencies
+RUN pip install --no-cache-dir flask PyMySQL psycopg2-binary
+
+# Copy the SQL executor script
+COPY scripts/sql-query-executor.py /app/app.py
+
+# Expose port
+EXPOSE 4000
+
+# Health check
+HEALTHCHECK --interval=10s --timeout=5s --retries=5 \
+  CMD python -c "import urllib.request; urllib.request.urlopen('http://localhost:4000/health').read()"
+
+# Run the app
+CMD ["python", "app.py"]

compose.yaml

@@ -134,6 +134,33 @@ services:
       retries: 5
       start_period: 10s
 
+  sql-executor:
+    build:
+      context: .
+      dockerfile: Dockerfile.sql-executor
+    restart: always
+    ports:
+      - "4000:4000"
+    environment:
+      - FREESCOUT_DB_HOST=10.136.40.104
+      - FREESCOUT_DB_PORT=3306
+      - FREESCOUT_DB_USER=freescout
+      - FREESCOUT_DB_PASSWORD=5N6fv4wIgsI6BZV
+      - FREESCOUT_DB_NAME=freescout
+      - POSTGRES_HOST=postgres
+      - POSTGRES_PORT=5432
+      - POSTGRES_USER=kb_user
+      - POSTGRES_PASSWORD=${POSTGRES_PASSWORD}
+      - POSTGRES_DB=n8n_kb
+    depends_on:
+      postgres:
+        condition: service_healthy
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:4000/health"]
+      interval: 10s
+      timeout: 5s
+      retries: 5
+
 volumes:
   n8n_data:
   traefik_data:

n8n-workflows/workflow-a-http.json

@@ -0,0 +1,274 @@
+{
+  "name": "Workflow A - Mail Processing (HTTP)",
+  "description": "Fetch unprocessed conversations from Freescout, analyze with AI, save suggestions",
+  "nodes": [
+    {
+      "id": "uuid-trigger-1",
+      "name": "Trigger",
+      "type": "n8n-nodes-base.cron",
+      "typeVersion": 1,
+      "position": [250, 200],
+      "parameters": {
+        "cronExpression": "*/5 * * * *"
+      }
+    },
+    {
+      "id": "uuid-get-conversations",
+      "name": "Get Unprocessed Conversations",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [450, 200],
+      "parameters": {
+        "url": "http://host.docker.internal:4000/query/freescout",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"query\":\"SELECT c.id, c.number, c.subject, c.customer_email, c.status, GROUP_CONCAT(t.body SEPARATOR '\\\\n') as threads_text FROM conversations c LEFT JOIN threads t ON c.id = t.conversation_id WHERE c.status = 1 AND c.id NOT IN (SELECT DISTINCT conversation_id FROM conversation_custom_field WHERE custom_field_id = 8) GROUP BY c.id LIMIT 20\"}"
+      }
+    },
+    {
+      "id": "uuid-split-results",
+      "name": "Split Results",
+      "type": "n8n-nodes-base.splitInBatches",
+      "typeVersion": 3,
+      "position": [650, 200],
+      "parameters": {
+        "batchSize": 1,
+        "options": {}
+      }
+    },
+    {
+      "id": "uuid-extract-data",
+      "name": "Extract Conversation Data",
+      "type": "n8n-nodes-base.set",
+      "typeVersion": 3,
+      "position": [850, 200],
+      "parameters": {
+        "options": {},
+        "assignments": {
+          "assignments": [
+            {
+              "name": "ticket_id",
+              "value": "={{ $json.id }}",
+              "type": "number"
+            },
+            {
+              "name": "ticket_number",
+              "value": "={{ $json.number }}",
+              "type": "number"
+            },
+            {
+              "name": "subject",
+              "value": "={{ $json.subject }}",
+              "type": "string"
+            },
+            {
+              "name": "problem_text",
+              "value": "={{ ($json.threads_text || 'No description provided').substring(0, 2000) }}",
+              "type": "string"
+            },
+            {
+              "name": "customer_email",
+              "value": "={{ $json.customer_email }}",
+              "type": "string"
+            }
+          ]
+        }
+      }
+    },
+    {
+      "id": "uuid-llm-analyze",
+      "name": "LiteLLM AI Analysis",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [1050, 200],
+      "parameters": {
+        "url": "http://llm.eks-ai.apps.asgard.eks-lnx.fft-it.de/v1/chat/completions",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"model\":\"gpt-3.5-turbo\",\"messages\":[{\"role\":\"system\",\"content\":\"Du bist ein IT-Support-Assistent. Analysiere das folgende IT-Support-Ticket und gib eine strukturierte JSON-Antwort mit folgenden Feldern: kategorie (z.B. Hardware, Software, Netzwerk, Zugriff), lösung_typ (BARAMUNDI_JOB, AUTOMATISCHE_ANTWORT, oder ESKALATION), vertrauen (Dezimal zwischen 0.0 und 1.0 - wie sicher bist du bei dieser Lösung), baramundi_job (Name des Jobs falls BARAMUNDI_JOB), antwort_text (Die Antwort an den Nutzer), begründung (Kurze Erklärung deiner Analyse)\"},{\"role\":\"user\",\"content\":\"Ticket-Nummer: {{$json.ticket_number}}\\nBetreff: {{$json.subject}}\\nProblembeschreibung:\\n{{$json.problem_text}}\\n\\nBitte antworte NUR mit gültiger JSON in dieser Struktur: {\\\"kategorie\\\": \\\"...\\\", \\\"lösung_typ\\\": \\\"...\\\", \\\"vertrauen\\\": 0.75, \\\"baramundi_job\\\": \\\"...\\\", \\\"antwort_text\\\": \\\"...\\\", \\\"begründung\\\": \\\"...\\\"}\"}],\"temperature\":0.7,\"max_tokens\":1000}"
+      }
+    },
+    {
+      "id": "uuid-parse-response",
+      "name": "Parse AI Response",
+      "type": "n8n-nodes-base.set",
+      "typeVersion": 3,
+      "position": [1250, 200],
+      "parameters": {
+        "options": {},
+        "assignments": {
+          "assignments": [
+            {
+              "name": "response_text",
+              "value": "={{ $json.choices[0].message.content }}",
+              "type": "string"
+            },
+            {
+              "name": "ai_response",
+              "value": "={{ JSON.parse($json.response_text) }}",
+              "type": "object"
+            }
+          ]
+        }
+      }
+    },
+    {
+      "id": "uuid-check-confidence",
+      "name": "Check Confidence >= 0.6",
+      "type": "n8n-nodes-base.if",
+      "typeVersion": 2,
+      "position": [1450, 200],
+      "parameters": {
+        "conditions": {
+          "options": {
+            "caseSensitive": true,
+            "extractValue": false
+          },
+          "combinator": "and",
+          "conditions": [
+            {
+              "id": "condition_1",
+              "leftValue": "={{ $json.ai_response.vertrauen }}",
+              "rightValue": 0.6,
+              "operator": {
+                "name": "filter.operator.gte",
+                "value": ">="
+              }
+            }
+          ]
+        }
+      }
+    },
+    {
+      "id": "uuid-save-to-db",
+      "name": "Save Suggestion to Freescout DB",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [1650, 100],
+      "parameters": {
+        "url": "http://host.docker.internal:4000/query/freescout",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"query\":\"INSERT INTO conversation_custom_field (conversation_id, custom_field_id, value) VALUES ({{$json.ticket_id}}, 6, '{{$json.ai_response | json.stringify}}') ON DUPLICATE KEY UPDATE value = VALUES(value); INSERT INTO conversation_custom_field (conversation_id, custom_field_id, value) VALUES ({{$json.ticket_id}}, 7, 'PENDING') ON DUPLICATE KEY UPDATE value = VALUES(value); INSERT INTO conversation_custom_field (conversation_id, custom_field_id, value) VALUES ({{$json.ticket_id}}, 8, '1') ON DUPLICATE KEY UPDATE value = VALUES(value);\"}"
+      }
+    },
+    {
+      "id": "uuid-no-action",
+      "name": "Skip - Low Confidence",
+      "type": "n8n-nodes-base.set",
+      "typeVersion": 3,
+      "position": [1650, 350],
+      "parameters": {
+        "options": {},
+        "assignments": {
+          "assignments": [
+            {
+              "name": "skipped",
+              "value": true,
+              "type": "boolean"
+            },
+            {
+              "name": "reason",
+              "value": "Confidence {{$json.ai_response.vertrauen}} < 0.6",
+              "type": "string"
+            }
+          ]
+        }
+      }
+    }
+  ],
+  "connections": {
+    "Trigger": {
+      "main": [
+        [
+          {
+            "node": "Get Unprocessed Conversations",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Get Unprocessed Conversations": {
+      "main": [
+        [
+          {
+            "node": "Split Results",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Split Results": {
+      "main": [
+        [
+          {
+            "node": "Extract Conversation Data",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Extract Conversation Data": {
+      "main": [
+        [
+          {
+            "node": "LiteLLM AI Analysis",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "LiteLLM AI Analysis": {
+      "main": [
+        [
+          {
+            "node": "Parse AI Response",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Parse AI Response": {
+      "main": [
+        [
+          {
+            "node": "Check Confidence >= 0.6",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Check Confidence >= 0.6": {
+      "main": [
+        [
+          {
+            "node": "Save Suggestion to Freescout DB",
+            "index": 0
+          }
+        ],
+        [
+          {
+            "node": "Skip - Low Confidence",
+            "index": 0
+          }
+        ]
+      ]
+    }
+  },
+  "active": false,
+  "settings": {
+    "errorHandler": "continueOnError"
+  }
+}

n8n-workflows/workflow-b-http.json

@@ -0,0 +1,337 @@
+{
+  "name": "Workflow B - Approval & Execution (HTTP)",
+  "description": "Poll for approved AI suggestions and execute them (Baramundi jobs or email replies)",
+  "nodes": [
+    {
+      "id": "uuid-trigger-b",
+      "name": "Trigger",
+      "type": "n8n-nodes-base.cron",
+      "typeVersion": 1,
+      "position": [250, 200],
+      "parameters": {
+        "cronExpression": "*/2 * * * *"
+      }
+    },
+    {
+      "id": "uuid-get-approved",
+      "name": "Get Approved Conversations",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [450, 200],
+      "parameters": {
+        "url": "http://host.docker.internal:4000/query/freescout",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"query\":\"SELECT c.id, c.number, c.subject, c.customer_email, ccf.value as ai_suggestion FROM conversations c JOIN conversation_custom_field ccf ON c.id = ccf.conversation_id WHERE ccf.custom_field_id = 7 AND ccf.value = 'APPROVED' LIMIT 10\"}"
+      }
+    },
+    {
+      "id": "uuid-split-approved",
+      "name": "Split Results",
+      "type": "n8n-nodes-base.splitInBatches",
+      "typeVersion": 3,
+      "position": [650, 200],
+      "parameters": {
+        "batchSize": 1,
+        "options": {}
+      }
+    },
+    {
+      "id": "uuid-extract-approved",
+      "name": "Extract & Parse Suggestion",
+      "type": "n8n-nodes-base.set",
+      "typeVersion": 3,
+      "position": [850, 200],
+      "parameters": {
+        "options": {},
+        "assignments": {
+          "assignments": [
+            {
+              "name": "ticket_id",
+              "value": "={{ $json.id }}",
+              "type": "number"
+            },
+            {
+              "name": "ticket_number",
+              "value": "={{ $json.number }}",
+              "type": "number"
+            },
+            {
+              "name": "subject",
+              "value": "={{ $json.subject }}",
+              "type": "string"
+            },
+            {
+              "name": "customer_email",
+              "value": "={{ $json.customer_email }}",
+              "type": "string"
+            },
+            {
+              "name": "ai_suggestion_raw",
+              "value": "={{ typeof $json.ai_suggestion === 'string' ? $json.ai_suggestion : JSON.stringify($json.ai_suggestion) }}",
+              "type": "string"
+            },
+            {
+              "name": "ai_suggestion",
+              "value": "={{ typeof $json.ai_suggestion === 'string' ? JSON.parse($json.ai_suggestion) : $json.ai_suggestion }}",
+              "type": "object"
+            },
+            {
+              "name": "solution_type",
+              "value": "={{ $json.ai_suggestion.lösung_typ || 'UNKNOWN' }}",
+              "type": "string"
+            }
+          ]
+        }
+      }
+    },
+    {
+      "id": "uuid-decide-solution",
+      "name": "Decide Solution Type",
+      "type": "n8n-nodes-base.switch",
+      "typeVersion": 1,
+      "position": [1050, 200],
+      "parameters": {
+        "options": [
+          {
+            "condition": "equal",
+            "value1": "={{ $json.solution_type }}",
+            "value2": "BARAMUNDI_JOB"
+          },
+          {
+            "condition": "equal",
+            "value1": "={{ $json.solution_type }}",
+            "value2": "AUTOMATISCHE_ANTWORT"
+          }
+        ]
+      }
+    },
+    {
+      "id": "uuid-execute-baramundi",
+      "name": "Execute Baramundi Job",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [1250, 50],
+      "parameters": {
+        "url": "https://baramundi-api.example.com/api/jobs",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json",
+          "Authorization": "Bearer YOUR_BARAMUNDI_TOKEN"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"job_name\":\"{{$json.ai_suggestion.baramundi_job}}\",\"ticket_id\":{{$json.ticket_id}},\"target_system\":\"IT\",\"description\":\"{{$json.subject}}\"}"
+      }
+    },
+    {
+      "id": "uuid-send-email",
+      "name": "Send Email Reply",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [1250, 150],
+      "parameters": {
+        "url": "http://host.docker.internal:4000/query/freescout",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"query\":\"INSERT INTO threads (conversation_id, customer_id, user_id, type, status, body, created_at, updated_at) VALUES ({{$json.ticket_id}}, (SELECT customer_id FROM conversations WHERE id = {{$json.ticket_id}} LIMIT 1), NULL, 'customer', 'active', '{{$json.ai_suggestion.antwort_text | replace(\\\"'\\\", \\\"''\\\")}}', NOW(), NOW())\"}"
+      }
+    },
+    {
+      "id": "uuid-mark-escalation",
+      "name": "Mark for Manual Review",
+      "type": "n8n-nodes-base.set",
+      "typeVersion": 3,
+      "position": [1250, 270],
+      "parameters": {
+        "options": {},
+        "assignments": {
+          "assignments": [
+            {
+              "name": "action",
+              "value": "Manual escalation required",
+              "type": "string"
+            }
+          ]
+        }
+      }
+    },
+    {
+      "id": "uuid-update-status",
+      "name": "Update Status to EXECUTED",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [1450, 200],
+      "parameters": {
+        "url": "http://host.docker.internal:4000/query/freescout",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"query\":\"UPDATE conversation_custom_field SET value = 'EXECUTED' WHERE conversation_id = {{$json.ticket_id}} AND custom_field_id = 7\"}"
+      }
+    },
+    {
+      "id": "uuid-trigger-workflow-c",
+      "name": "Trigger Workflow C (KB Update)",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [1650, 200],
+      "parameters": {
+        "url": "https://n8n.fft-it.de/webhook/workflow-c",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"ticket_id\":{{$json.ticket_id}},\"subject\":\"{{$json.subject}}\",\"problem\":\"{{$json.subject}}\",\"solution\":\"{{$json.ai_suggestion.antwort_text}}\",\"category\":\"{{$json.ai_suggestion.kategorie}}\",\"solution_type\":\"{{$json.solution_type}}\"}"
+      }
+    },
+    {
+      "id": "uuid-log-audit",
+      "name": "Log to PostgreSQL",
+      "type": "n8n-nodes-base.httpRequest",
+      "typeVersion": 4,
+      "position": [1850, 200],
+      "parameters": {
+        "url": "http://host.docker.internal:4000/query/audit",
+        "method": "POST",
+        "headers": {
+          "Content-Type": "application/json"
+        },
+        "sendBody": true,
+        "specifyBody": "json",
+        "jsonBody": "{\"query\":\"INSERT INTO workflow_executions (workflow_name, ticket_id, status, execution_time_ms, created_at) VALUES ('Workflow B - Approval Execution', {{$json.ticket_id}}, 'SUCCESS', 0, NOW())\"}"
+      }
+    }
+  ],
+  "connections": {
+    "Trigger": {
+      "main": [
+        [
+          {
+            "node": "Get Approved Conversations",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Get Approved Conversations": {
+      "main": [
+        [
+          {
+            "node": "Split Results",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Split Results": {
+      "main": [
+        [
+          {
+            "node": "Extract & Parse Suggestion",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Extract & Parse Suggestion": {
+      "main": [
+        [
+          {
+            "node": "Decide Solution Type",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Decide Solution Type": {
+      "main": [
+        [
+          {
+            "node": "Execute Baramundi Job",
+            "index": 0
+          }
+        ],
+        [
+          {
+            "node": "Send Email Reply",
+            "index": 0
+          }
+        ],
+        [
+          {
+            "node": "Mark for Manual Review",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Execute Baramundi Job": {
+      "main": [
+        [
+          {
+            "node": "Update Status to EXECUTED",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Send Email Reply": {
+      "main": [
+        [
+          {
+            "node": "Update Status to EXECUTED",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Mark for Manual Review": {
+      "main": [
+        [
+          {
+            "node": "Update Status to EXECUTED",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Update Status to EXECUTED": {
+      "main": [
+        [
+          {
+            "node": "Trigger Workflow C (KB Update)",
+            "index": 0
+          }
+        ]
+      ]
+    },
+    "Trigger Workflow C (KB Update)": {
+      "main": [
+        [
+          {
+            "node": "Log to PostgreSQL",
+            "index": 0
+          }
+        ]
+      ]
+    }
+  },
+  "active": false,
+  "settings": {
+    "errorHandler": "continueOnError"
+  }
+}

scripts/sql-query-executor.py

@@ -0,0 +1,197 @@
+#!/usr/bin/env python3
+"""
+Simple HTTP Server for executing SQL queries
+Used by n8n workflows to avoid needing specialized database nodes
+"""
+
+from flask import Flask, request, jsonify
+import pymysql
+import psycopg2
+import logging
+import os
+
+app = Flask(__name__)
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+# Database configuration
+FREESCOUT_DB_CONFIG = {
+    'host': os.getenv('FREESCOUT_DB_HOST', '10.136.40.104'),
+    'port': int(os.getenv('FREESCOUT_DB_PORT', 3306)),
+    'user': os.getenv('FREESCOUT_DB_USER', 'freescout'),
+    'password': os.getenv('FREESCOUT_DB_PASSWORD', '5N6fv4wIgsI6BZV'),
+    'database': os.getenv('FREESCOUT_DB_NAME', 'freescout'),
+    'charset': 'utf8mb4',
+    'autocommit': True,
+}
+
+POSTGRES_AUDIT_CONFIG = {
+    'host': os.getenv('POSTGRES_HOST', 'postgres'),
+    'port': int(os.getenv('POSTGRES_PORT', 5432)),
+    'user': os.getenv('POSTGRES_USER', 'kb_user'),
+    'password': os.getenv('POSTGRES_PASSWORD', 'change_me_securely'),
+    'database': os.getenv('POSTGRES_DB', 'n8n_kb'),
+}
+
+
+def execute_query(db_type, query):
+    """
+    Execute a SQL query and return results
+    db_type: 'freescout' or 'audit'
+    """
+    connection = None
+    cursor = None
+
+    try:
+        if db_type == 'freescout':
+            connection = pymysql.connect(**FREESCOUT_DB_CONFIG)
+            cursor = connection.cursor(pymysql.cursors.DictCursor)
+        elif db_type == 'audit':
+            connection = psycopg2.connect(
+                host=POSTGRES_AUDIT_CONFIG['host'],
+                port=POSTGRES_AUDIT_CONFIG['port'],
+                user=POSTGRES_AUDIT_CONFIG['user'],
+                password=POSTGRES_AUDIT_CONFIG['password'],
+                database=POSTGRES_AUDIT_CONFIG['database']
+            )
+            cursor = connection.cursor()
+        else:
+            return None, "Invalid database type"
+
+        logger.info(f"Executing {db_type} query: {query[:100]}...")
+        cursor.execute(query)
+
+        if query.strip().upper().startswith('SELECT'):
+            # Fetch results for SELECT queries
+            if db_type == 'freescout':
+                results = cursor.fetchall()
+            else:
+                # PostgreSQL: convert to list of dicts
+                columns = [desc[0] for desc in cursor.description]
+                results = [dict(zip(columns, row)) for row in cursor.fetchall()]
+            return results, None
+        else:
+            # For INSERT/UPDATE/DELETE
+            connection.commit()
+            return {'affected_rows': cursor.rowcount}, None
+
+    except pymysql.Error as e:
+        logger.error(f"Database error: {e}")
+        return None, str(e)
+    except Exception as e:
+        logger.error(f"Error: {e}")
+        return None, str(e)
+    finally:
+        if cursor:
+            cursor.close()
+        if connection:
+            try:
+                connection.close()
+            except:
+                pass
+
+
+@app.route('/health', methods=['GET'])
+def health():
+    """Health check endpoint"""
+    return jsonify({'status': 'ok', 'service': 'sql-executor'}), 200
+
+
+@app.route('/query', methods=['POST'])
+def query():
+    """
+    Execute a SQL query
+
+    Request body:
+    {
+        "db_type": "freescout" or "audit",
+        "query": "SELECT * FROM conversations LIMIT 10"
+    }
+    """
+    try:
+        data = request.get_json()
+
+        if not data or 'query' not in data:
+            return jsonify({'error': 'Missing query parameter'}), 400
+
+        db_type = data.get('db_type', 'freescout')
+        query_str = data.get('query')
+
+        results, error = execute_query(db_type, query_str)
+
+        if error:
+            logger.error(f"Query failed: {error}")
+            return jsonify({'error': error, 'success': False}), 500
+
+        return jsonify({
+            'success': True,
+            'data': results,
+            'count': len(results) if isinstance(results, list) else 1
+        }), 200
+
+    except Exception as e:
+        logger.error(f"Error: {e}")
+        return jsonify({'error': str(e), 'success': False}), 500
+
+
+@app.route('/query/freescout', methods=['POST'])
+def query_freescout():
+    """Execute query on Freescout database"""
+    try:
+        data = request.get_json()
+        if not data or 'query' not in data:
+            return jsonify({'error': 'Missing query parameter', 'success': False}), 400
+
+        query_str = data.get('query')
+        results, error = execute_query('freescout', query_str)
+
+        if error:
+            logger.error(f"Query failed: {error}")
+            return jsonify({'error': error, 'success': False}), 500
+
+        return jsonify({
+            'success': True,
+            'data': results,
+            'count': len(results) if isinstance(results, list) else 1
+        }), 200
+    except Exception as e:
+        logger.error(f"Error: {e}")
+        return jsonify({'error': str(e), 'success': False}), 500
+
+
+@app.route('/query/audit', methods=['POST'])
+def query_audit():
+    """Execute query on Audit (PostgreSQL) database"""
+    try:
+        data = request.get_json()
+        if not data or 'query' not in data:
+            return jsonify({'error': 'Missing query parameter', 'success': False}), 400
+
+        query_str = data.get('query')
+        results, error = execute_query('audit', query_str)
+
+        if error:
+            logger.error(f"Query failed: {error}")
+            return jsonify({'error': error, 'success': False}), 500
+
+        return jsonify({
+            'success': True,
+            'data': results,
+            'count': len(results) if isinstance(results, list) else 1
+        }), 200
+    except Exception as e:
+        logger.error(f"Error: {e}")
+        return jsonify({'error': str(e), 'success': False}), 500
+
+
+if __name__ == '__main__':
+    # Test connection on startup
+    logger.info("Testing Freescout database connection...")
+    results, error = execute_query('freescout', 'SELECT 1')
+    if error:
+        logger.warning(f"Freescout DB connection test failed: {error} (will retry during runtime)")
+    else:
+        logger.info(f"✓ Connected to Freescout DB")
+
+    logger.info("Starting SQL Query Executor on 0.0.0.0:4000")
+    app.run(host='0.0.0.0', port=4000, debug=False, threaded=True)