upphandlat-mcp
If you are the rightful owner of upphandlat-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
This MCP server provides access to Swedish public procurement data, enabling analysis of procurements, their value, and supplier details.
The Upphandlat MCP Server is a specialized tool designed to facilitate the analysis of public procurement data from Sweden. It sources its data from the Swedish National Procurement Agency, known as Upphandlingsmyndigheten. The server is built to handle large datasets efficiently using Polars, a high-performance data manipulation library. By leveraging the Model Context Protocol (MCP), it offers a robust interface for interacting with procurement data, allowing users to perform complex data operations such as aggregations, filtering, and fuzzy searches. This server is particularly useful for researchers, policymakers, and businesses interested in understanding procurement trends and supplier dynamics in Sweden. It provides a structured approach to accessing and analyzing data, making it easier to derive insights and make informed decisions.
Features
- Load multiple CSV datasets from specified URLs on server startup.
- Expose CSV schema and column information for easy data exploration.
- Provide powerful aggregation capabilities, including grouping, filtering, and calculations.
- Conduct fuzzy searches within text columns to find relevant information.
- Add configurable summary rows to aggregation results for enhanced data analysis.
Tools
list_available_dataframes
Lists all loaded datasets with their names and descriptions.
list_columns
Returns column names for a specified dataset.
get_schema
Returns the schema (column names and data types) for a dataset.
get_distinct_column_values
Retrieves unique values from a column.
fuzzy_search_column_values
Performs fuzzy matching in a text column.
aggregate_data
The main tool for filtering, grouping, aggregating, and calculating new fields from the data.