Give Claude Code live access to institutional data.
After one setup command, Claude Code treats Apiar Data as a live tool it can call on its own. Give it a research task — a sector analysis, a country overview, a macro comparison — and it will search, retrieve, and use the data without you specifying each query.
Run this command to install
Run it once in your terminal. From that point on, Claude Code — including any sub-agents it spawns — can search institutional data autonomously during a task.
claude mcp add --transport http ApiarData https://search.apiardata.com/mcpThree-step setup
Run the command above in your terminal. Claude Code registers the server immediately — no restart needed.
Start any task that involves data — a market brief, a sector comparison, an economic overview. Claude will call the search tool automatically when it needs figures.
Run /mcp to confirm the server is listed, or ask Claude directly whether it has the Apiar Data tool available.
Optional setup snippets
What the command writes
Optional reference. This is the entry added to ~/.claude/settings.json — useful if you manage config files manually or want to replicate the setup across machines.
{
"mcpServers": {
"ApiarData": {
"type": "http",
"url": "https://search.apiardata.com/mcp"
}
}
}A few practical notes
- The server is designed for agentic use: Claude Code decides when to call it. You can still ask for a specific dataset explicitly, but in most research tasks Claude will search without being prompted.
- Sub-agents and background agents spawned by Claude Code inherit the same tool access, so multi-step research pipelines can pull live data end-to-end without manual intervention.
- Delegate a full research task — sector analysis, country comparison, macro overview — and Claude retrieves the figures, chains queries across multiple datasets, and uses them directly in its response.
- Every result links to the source institution — ECB, Eurostat, IMF, national statistical offices — so you can trace any figure back to its origin.