SuperAPI AI Directory

apiverket-mcp

Query Swedish public data through Apiverket, including company data, SCB statistics, weather, transport and more.

Open website

About this tool

Apiverket MCP Server is an MCP (Model Context Protocol) server that provides AI assistants with the ability to access Swedish public data through the Apiverket API. It works with Claude Code, Claude Desktop, Cursor, VS Code, Gemini CLI-style MCP clients, and other MCP-compatible tools.

Purpose The server is designed to enable AI agents to query Swedish government public data, covering company data, SCB statistics, weather, traffic, and many other domains. Agents should discover and query data through a structured workflow rather than guessing API paths.

Main capabilities

  • Provides the govdata_discover tool for searching the built-in endpoint catalog based on a user topic, without making API calls.
  • Provides the govdata_account tool for querying quota, tier, and production readiness status.
  • Provides the govdata_query tool for calling the Apiverket API using only supported endpoints returned by discovery.
  • The discovery feature includes workflow series for traffic, weather, geography, municipalities, employment, government, culture, education, tourism, environment, and company data, explaining search/list/detail patterns, required codes, coordinates, date formats, pagination, and common recovery steps.
  • Structured API errors are converted into recovery guidance for agents.
  • Large responses are truncated and pagination guidance is provided.

Usage Agents should follow these steps:

  1. Discover: Call govdata_discover with the user topic instead of guessing paths.
  2. Check context: Call govdata_account for quota, tier, and production readiness questions.
  3. Query: Call govdata_query using only supported endpoints returned by discovery.
  4. Recover: If a request fails, use the structured recovery guidance before retrying.

Company data usage:

  • Use govdata_discover(query: "company") to find supported company endpoints.
  • When the user only has a company name or is unsure of the organization number, use /v1/companies/search.
  • Cache the returned org_number.
  • For repeated enrichment and automation, use /v1/companies/{orgNumber}.
  • If companysearch returns 429, stop retrying before reset_at; use lookup when the organization number is known.

Keys and limits sk_test_demo is used for setup and safe sandbox responses, suitable for checking whether the MCP client is connected correctly, but production workflows should use a live key. A free live key is suitable for exploration; upgrade when daily API or company search limits hinder actual workflows. Company search and live company lookup have separate daily quota contexts by tier. govdata_account shows the configured key tier, remaining company search quota, remaining company lookup quota, reset time, and available upgrade tiers. Apiverket does not expose company board, executives, owners, or UBO sub-resource paths through the company API.

Building from source The MCP server runs locally via stdio. An Apiverket API key is required, or use sk_test_demo for sandbox setup.

Use cases Suitable for AI assistant workflows that need to query Swedish public data, including company information queries, statistics retrieval, weather and traffic information, and more.

Limitations

  • An Apiverket API key is required, or use sk_test_demo for sandbox testing.
  • The company API does not provide board, executives, owners, or UBO sub-resource paths.
  • Free live keys have daily API or company search limits.
  • Company search and company lookup have independent daily quotas.
  • Large responses are truncated.
  • Specific installation commands and configuration details are not provided at this time.