SafetyCulture MCP Server

What is SafetyCulture MCP Server?

SafetyCulture MCP Server is a Model Context Protocol server that connects to the SafetyCulture API. It lets users ask natural language questions about their SafetyCulture data after providing an API key, enabling inspection analysis and trend reporting.

How to use SafetyCulture MCP Server?

Clone the repository, install dependencies with pip install -r requirements.txt, copy example.env to .env and configure your SafetyCulture API key. Run the server using run_server.bat (reads from .env) or run_with_key.bat YOUR_API_KEY. For use with Claude for Desktop, add the server to the MCP configuration file (claude_desktop_config.json) pointing to src/main.py.

Key features of SafetyCulture MCP Server

• Query SafetyCulture data using natural language
• Analyze inspection data and trends over time
• Compare safety metrics across time periods and categories
• Visualize inspection trends over time
• Authenticate with the SafetyCulture API via API key
• Retrieve and filter actions by status and priority

Use cases of SafetyCulture MCP Server

• Count inspections at a specific site over the last 3 months
• Compare injury reports between two time periods
• Track action items filtered by status or priority
• Identify safety trend patterns across categories

FAQ from SafetyCulture MCP Server

What authentication is required?

You must provide a valid SafetyCulture API key, either stored in a .env file or passed directly as a command-line argument.

What data does the server access?

It uses the SafetyCulture Feed API to retrieve inspections and actions, including their metadata, status, priority, and time periods.

Does the server have any dependencies?

Yes, it requires Python and the packages listed in requirements.txt. No additional database or cloud service is needed beyond the SafetyCulture account and API key.

How is the server run with Claude for Desktop?

Configure the mcpServers section in Claude Desktop’s configuration file with the command python and the full path to src/main.py as arguments.

What tools are available?

Tools include authenticate, get_inspections, get_inspection_trends, compare_injury_reports, get_actions, and get_action_details.