qlik-sense-mcp-server 1.1.1

MCP Server for Qlik Sense Enterprise APIs

Qlik Sense MCP Server

Model Context Protocol (MCP) server for integration with Qlik Sense Enterprise APIs. Provides unified interface for Repository API and Engine API operations through MCP protocol.

Table of Contents

• Overview
• Features
• Installation
• Configuration
• Usage
• API Reference
• Architecture
• Development
• Troubleshooting
• License

Overview

Qlik Sense MCP Server bridges Qlik Sense Enterprise with systems supporting Model Context Protocol. Server provides 6 essential tools for application metadata retrieval and data analysis operations.

Key Features

• Unified API: Single interface for Qlik Sense Repository and Engine APIs
• Security: Certificate-based authentication support
• Performance: Optimized queries and direct API access
• Analytics: Advanced data analysis and hypercube creation
• Metadata: Comprehensive application and field information

Features

Available Tools

Tool	Description	API	Status
get_apps	Get comprehensive list of applications with metadata	Repository	✅
get_app_details	Get detailed application analysis including data model	Engine	✅
engine_get_script	Extract load script from application	Engine	✅
engine_get_field_values	Get field values with frequency information	Engine	✅
engine_get_field_statistics	Get comprehensive field statistics	Engine	✅
engine_create_hypercube	Create hypercube for data analysis	Engine	✅

Installation

Quick Start with uvx (Recommended)

The easiest way to use Qlik Sense MCP Server is with uvx:

uvx qlik-sense-mcp-server

This command will automatically install and run the latest version without affecting your system Python environment.

Alternative Installation Methods

From PyPI

pip install qlik-sense-mcp-server

From Source (Development)

git clone https://github.com/bintocher/qlik-sense-mcp.git
cd qlik-sense-mcp
make dev

System Requirements

• Python 3.12+
• Qlik Sense Enterprise
• Valid certificates for authentication
• Network access to Qlik Sense server (ports 4242, 4747)

Setup

1. Setup certificates

mkdir certs
# Copy your Qlik Sense certificates to certs/ directory

2. Create configuration

cp .env.example .env
# Edit .env with your settings

Configuration

Environment Variables (.env)

# Server connection
QLIK_SERVER_URL=https://your-qlik-server.company.com
QLIK_USER_DIRECTORY=COMPANY
QLIK_USER_ID=your-username

# Certificate paths (absolute paths)
QLIK_CLIENT_CERT_PATH=/path/to/certs/client.pem
QLIK_CLIENT_KEY_PATH=/path/to/certs/client_key.pem
QLIK_CA_CERT_PATH=/path/to/certs/root.pem

# API ports (standard Qlik Sense ports)
QLIK_REPOSITORY_PORT=4242
QLIK_ENGINE_PORT=4747

# SSL settings
QLIK_VERIFY_SSL=false

MCP Configuration

Create mcp.json file for MCP client integration:

{
  "mcpServers": {
    "qlik-sense": {
      "command": "uvx",
      "args": ["qlik-sense-mcp-server"],
      "env": {
        "QLIK_SERVER_URL": "https://your-qlik-server.company.com",
        "QLIK_USER_DIRECTORY": "COMPANY",
        "QLIK_USER_ID": "your-username",
        "QLIK_CLIENT_CERT_PATH": "/path/to/certs/client.pem",
        "QLIK_CLIENT_KEY_PATH": "/path/to/certs/client_key.pem",
        "QLIK_CA_CERT_PATH": "/path/to/certs/root.pem",
        "QLIK_REPOSITORY_PORT": "4242",
        "QLIK_ENGINE_PORT": "4747",
        "QLIK_VERIFY_SSL": "false"
      },
      "disabled": false,
      "autoApprove": [
        "get_apps",
        "get_app_details",
        "engine_get_script",
        "engine_get_field_values",
        "engine_get_field_statistics",
        "engine_create_hypercube"
      ]
    }
  }
}

Usage

Start Server

# Using uvx (recommended)
uvx qlik-sense-mcp-server

# Using installed package
qlik-sense-mcp-server

# From source (development)
python -m qlik_sense_mcp_server.server

Example Operations

Get Applications List

# Via MCP client - get first 50 apps (default)
result = mcp_client.call_tool("get_apps")
print(f"Showing {result['pagination']['returned']} of {result['pagination']['total_found']} apps")

# Search for specific apps
result = mcp_client.call_tool("get_apps", {
    "name_filter": "Sales",
    "limit": 10
})

# Get more apps (pagination)
result = mcp_client.call_tool("get_apps", {
    "offset": 50,
    "limit": 50
})

Analyze Application

# Get comprehensive app analysis
result = mcp_client.call_tool("get_app_details", {
    "app_id": "your-app-id"
})
print(f"App has {len(result['data_model']['tables'])} tables")

Create Data Analysis Hypercube

# Create hypercube for sales analysis
result = mcp_client.call_tool("engine_create_hypercube", {
    "app_id": "your-app-id",
    "dimensions": ["Region", "Product"],
    "measures": ["Sum(Sales)", "Count(Orders)"],
    "max_rows": 1000
})

Get Field Statistics

# Get detailed field statistics
result = mcp_client.call_tool("engine_get_field_statistics", {
    "app_id": "your-app-id",
    "field_name": "Sales"
})
print(f"Average: {result['avg_value']['numeric']}")

API Reference

get_apps

Retrieves comprehensive list of Qlik Sense applications with metadata, pagination and filtering support.

Parameters:

• limit (optional): Maximum number of apps to return (default: 50, max: 1000)
• offset (optional): Number of apps to skip for pagination (default: 0)
• name_filter (optional): Filter apps by name (case-insensitive partial match)
• app_id_filter (optional): Filter by specific app ID/GUID
• include_unpublished (optional): Include unpublished apps (default: true)

Returns: Object containing paginated apps, streams, and pagination metadata

Example (default - first 50 apps):

{
  "apps": [...],
  "streams": [...],
  "pagination": {
    "limit": 50,
    "offset": 0,
    "returned": 50,
    "total_found": 1598,
    "has_more": true,
    "next_offset": 50
  },
  "filters": {
    "name_filter": null,
    "app_id_filter": null,
    "include_unpublished": true
  },
  "summary": {
    "total_apps": 1598,
    "published_apps": 857,
    "private_apps": 741,
    "total_streams": 40,
    "showing": "1-50 of 1598"
  }
}

Example (with name filter):

# Search for apps containing "dashboard"
result = mcp_client.call_tool("get_apps", {
    "name_filter": "dashboard",
    "limit": 10
})

# Get specific app by ID
result = mcp_client.call_tool("get_apps", {
    "app_id_filter": "e2958865-2aed-4f8a-b3c7-20e6f21d275c"
})

# Get next page of results
result = mcp_client.call_tool("get_apps", {
    "limit": 50,
    "offset": 50
})

get_app_details

Gets comprehensive application analysis including data model, object counts, and metadata.

Parameters:

• app_id (required): Application identifier

Returns: Detailed application object with data model structure

Example:

{
  "app_metadata": {...},
  "data_model": {
    "tables": [...],
    "total_tables": 2,
    "total_fields": 45
  },
  "object_counts": {...}
}

engine_get_script

Retrieves load script from application.

Parameters:

• app_id (required): Application identifier

Returns: Object containing script text and metadata

Example:

{
  "qScript": "SET DateFormat='DD.MM.YYYY';\n...",
  "app_id": "app-id",
  "script_length": 2830
}

engine_get_field_values

Gets field values with frequency information.

Parameters:

• app_id (required): Application identifier
• field_name (required): Field name
• max_values (optional): Maximum values to return (default: 100)
• include_frequency (optional): Include frequency data (default: true)

Returns: Field values with statistics and dimension information

Example:

{
  "field_name": "state_name",
  "values": [
    {"value": "Найден, жив", "state": "O"},
    {"value": "Найден, погиб", "state": "O"}
  ],
  "total_values": 10
}

engine_get_field_statistics

Retrieves comprehensive field statistics.

Parameters:

• app_id (required): Application identifier
• field_name (required): Field name

Returns: Statistical analysis including min, max, average, median, mode, standard deviation

Example:

{
  "field_name": "age",
  "min_value": {"numeric": 0},
  "max_value": {"numeric": 2023},
  "avg_value": {"numeric": 40.98},
  "median_value": {"numeric": 38},
  "std_deviation": {"numeric": 24.88}
}

engine_create_hypercube

Creates hypercube for data analysis.

Parameters:

• app_id (required): Application identifier
• dimensions (required): Array of dimension fields
• measures (required): Array of measure expressions
• max_rows (optional): Maximum rows to return (default: 1000)

Returns: Hypercube data with dimensions, measures, and total statistics

Example:

{
  "hypercube_data": {
    "qDimensionInfo": [...],
    "qMeasureInfo": [...],
    "qDataPages": [...]
  },
  "total_rows": 30,
  "total_columns": 4
}

Architecture

Project Structure

qlik-sense-mcp/
├── qlik_sense_mcp_server/
│   ├── __init__.py
│   ├── server.py          # Main MCP server
│   ├── config.py          # Configuration management
│   ├── repository_api.py  # Repository API client (HTTP)
│   ├── engine_api.py      # Engine API client (WebSocket)
│   └── utils.py           # Utility functions
├── certs/                 # Certificates (git ignored)
│   ├── client.pem
│   ├── client_key.pem
│   └── root.pem
├── .env.example          # Configuration template
├── mcp.json.example      # MCP configuration template
├── pyproject.toml        # Project dependencies
└── README.md

System Components

QlikSenseMCPServer

Main server class handling MCP protocol operations, tool registration, and request routing.

QlikRepositoryAPI

HTTP client for Repository API operations including application metadata and administrative functions.

QlikEngineAPI

WebSocket client for Engine API operations including data extraction, analytics, and hypercube creation.

QlikSenseConfig

Configuration management class handling environment variables, certificate paths, and connection settings.

Development

Development Environment Setup

# Setup development environment
make dev

# Show all available commands
make help

# Build package
make build

Version Management

# Bump patch version and create PR
make version-patch

# Bump minor version and create PR
make version-minor

# Bump major version and create PR
make version-major

Adding New Tools

1. Add tool definition in server.py

# In tools_list
Tool(name="new_tool", description="Tool description", inputSchema={...})

2. Add handler in server.py

# In handle_call_tool()
elif name == "new_tool":
    result = await asyncio.to_thread(self.api_client.new_method, arguments)
    return [TextContent(type="text", text=json.dumps(result, indent=2))]

3. Implement method in API client

# In repository_api.py or engine_api.py
def new_method(self, param: str) -> Dict[str, Any]:
    """Method implementation."""
    return result

Troubleshooting

Common Issues

Certificate Errors

SSL: CERTIFICATE_VERIFY_FAILED

Solution:

• Verify certificate paths in .env
• Check certificate expiration
• Set QLIK_VERIFY_SSL=false for testing

Connection Errors

ConnectionError: Failed to connect to Engine API

Solution:

• Verify port 4747 accessibility
• Check server URL correctness
• Verify firewall settings

Authentication Errors

401 Unauthorized

Solution:

• Verify QLIK_USER_DIRECTORY and QLIK_USER_ID
• Check user exists in Qlik Sense
• Verify user permissions

Diagnostics

Test Configuration

python -c "
from qlik_sense_mcp_server.config import QlikSenseConfig
config = QlikSenseConfig.from_env()
print('Config valid:', config and hasattr(config, 'server_url'))
print('Server URL:', getattr(config, 'server_url', 'Not set'))
"

Test Repository API

python -c "
from qlik_sense_mcp_server.server import QlikSenseMCPServer
server = QlikSenseMCPServer()
print('Server initialized:', server.config_valid)
"

Performance

Optimization Recommendations

1. Use filters to limit data volume
2. Limit result size with max_rows parameter
3. Use Repository API for metadata (faster than Engine API)

Benchmarks

Operation	Average Time	Recommendations
get_apps	0.5s	Use filters
get_app_details	2-5s	Analyze specific apps
engine_create_hypercube	1-10s	Limit dimensions and measures
engine_get_field_statistics	0.5-2s	Use for numeric fields

Security

Recommendations

1. Store certificates securely - exclude from git
2. Use environment variables for sensitive data
3. Limit user permissions in Qlik Sense
4. Update certificates regularly
5. Monitor API access

Access Control

Create user in QMC with minimal required permissions:

• Read applications
• Access Engine API
• View data (if needed for analysis)

License

MIT License

Copyright (c) 2025 Stanislav Chernov

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

---

Project Status: Production Ready | 6/6 Tools Working | v1.1.1

Installation: uvx qlik-sense-mcp-server