ausecon-mcp-server 1.7.0

MCP server for Australian economic data from the ABS, RBA, and APRA.

Australian Economic Data (ABS, RBA & APRA) MCP Server

ausecon-mcp-server is a Python Model Context Protocol (MCP) server for structured Australian macroeconomic and financial data from the Australian Bureau of Statistics (ABS), the Reserve Bank of Australia (RBA), and the Australian Prudential Regulation Authority (APRA).

Version 1.6.0 is the current release line. Transport support is stdio plus Streamable HTTP. The server exposes fourteen read-only MCP tools, four read-only MCP resources, eight prompt templates, 70 curated analyst-facing economic and financial concepts through get_economic_series, and nine transparent derived indicators through get_derived_series.

Documentation

Full user and maintainer documentation is published at auseconmcp.com.

Useful links:

• Getting started
• Client setup
• Tool reference
• Semantic concepts
• Response schema
• Roadmap
• Changelog

Install

The package is published to PyPI and is intended to be launched by an MCP client on demand via uvx:

uvx ausecon-mcp-server

The server speaks MCP over standard input/output. When launched manually, it waits for a client to connect.

Client Setup

Claude Desktop:

{
  "mcpServers": {
    "ausecon": {
      "command": "uvx",
      "args": ["ausecon-mcp-server"]
    }
  }
}

Claude Code:

claude mcp add --transport stdio ausecon -- uvx ausecon-mcp-server

Codex:

codex mcp add ausecon -- uvx ausecon-mcp-server

Smithery:

This repository also includes smithery.yaml and Dockerfile.smithery for hosted Smithery custom container deployment over MCP Streamable HTTP at /mcp. The hosted HTTP entrypoint is ausecon-mcp-http; local users should keep using the stdio command above unless they are testing a container deployment. Maintainers can follow the deployment checklist in docs/smithery-deployment.md.

Basic Workflow

For normal economic concepts, discover the supported concept first:

list_economic_concepts(query="cash rate")

Then retrieve the resolved series:

get_economic_series(
  concept="cash_rate_target",
  start="2020-01-01"
)

For transparent formula-based indicators, call the derived retrieval surface directly:

get_derived_series(concept="real_cash_rate", last_n=12)

For exact source-native control, use search_datasets, list_catalogue, get_abs_dataset_structure, get_abs_data, get_rba_table, and get_apra_data.

For quick-turnaround analysis, use get_latest_observations, get_top_observations, describe_dataset, and list_release_events as additive convenience tools. These wrappers keep native ABS, RBA, and APRA identifiers visible, including source controls, safe convenience calls, and APRA expected-release cadence estimates.

Prompting an AI Agent

When connected to an LLM or AI agent, users can ask in ordinary economic language and the agent can translate that request into MCP tool calls. For example, a quarterly real GDP growth prompt usually maps to list_economic_concepts(query="quarterly real GDP growth") and then get_economic_series(concept="gdp_growth", last_n=40). The server returns structured metadata, series, and observations; the LLM client decides when to call tools and how to summarise the results. See the full AI-agent prompting guide.

Development

Python 3.12 is recommended for local development. The package metadata and CI matrix support Python 3.10+.

uv sync --python 3.12 --extra dev
uv run pytest
uv run ruff check src tests scripts

Repository

• Repository: github.com/AnthonyPuggs/ausecon-mcp-server
• Issues: GitHub Issues
• Licence: MIT