Rba MCP Server

@pipeworx/rba

Reserve Bank of Australia statistics — cash rate target, AUD exchange rates, and any of the 71 published statistical tables (interest rates, bond yields, CPI, monetary aggregates, and more), scraped keyless from RBA's own CSV endpoints.

Part of Pipeworx — an MCP gateway connecting AI agents to 1866+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Tools

• rba_cash_rate(recent?) — the RBA's official cash rate target (AU's monetary-policy rate) plus recent monthly history.
• rba_exchange_rates(currency?, recent?) — latest AUD exchange rates against major currencies, or one pair's recent history.
• rba_series(series?, table?, series_id?, recent?) — any RBA statistical series. Preferred usage: pass series as a NAME or description (e.g. "90-day bank bill rate", "3-year government bond yield") and it resolves the (table, series_id) pair by fuzzy-matching series titles/descriptions — the interest-rate and bond-yield tables (F1, F1.1, F2, F2.1, F16) are searched first, then the broader catalog if nothing scores there. Returns resolved_from/resolved_to on a clean resolution, or an error: "ambiguous" with candidates when the name doesn't disambiguate on its own (never invents an id). Pass table + series_id directly for the legacy exact-id path.
• rba_list_series(query?, keyword?, table?) — discovery/directory tool. query searches series NAMES/descriptions directly (same engine rba_series uses to resolve names) and returns ranked (table, series_id) matches in one call. keyword searches table titles/categories instead and returns every series in the matched tables. table lists all series in one table. No arguments browses every table by category.

Auth

Keyless. RBA publishes every statistical table as CSV with no authentication or rate limit documented.

Data sources

• https://www.rba.gov.au/statistics/tables/csv/<table>-data.csv — one CSV per table id (e.g. f1.1, f2.1, g1). Row layout is fixed: a title line, then labelled header rows (Title, Description, Frequency, Type, Units, Source, Publication date, Series ID), then data rows keyed by date. Columns 1..N are series, matched by position across the header rows.
• https://www.rba.gov.au/statistics/tables/ — the index page RBA_TABLES in src/index.ts was scraped from (2026-08-28). No official endpoint lists every table + series in one call; regenerate that scrape if RBA adds/renames a table (rare).

Gotchas

• Dates use two formats across tables: DD/MM/YYYY (e.g. f1.1) and DD-Mon-YYYY (e.g. f11.1 exchange rates) — both normalized to YYYY-MM-DD by toIso().
• RBA republishes several series in both a daily and a monthly table with the same or near-identical name (e.g. "Cash Rate Target" in both f1 and f1.1; "EOD 3-month BABs/NCDs" in f1 vs "3-month BABs/NCDs" in f1.1). Name resolution collapses these via dedupeMatches() and prefers the monthly/interpolated table — otherwise every named query looks falsely ambiguous against its own daily twin.
• Day-count finance shorthand ("90-day", "180-day") doesn't literally appear in RBA's own series titles — RBA calls them "1/3/6-month BABs/NCDs". expandDaySynonyms() converts common day counts (30/90/180/365) to their month equivalent before scoring, so "90-day bank bill rate" still finds FIRMMBAB90 ("3-month BABs/NCDs").

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "rba": {
      "url": "https://gateway.pipeworx.io/rba/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/rba/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1866+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

curl -X POST https://gateway.pipeworx.io/v1/tools/rba_cash_rate \
  -H 'Content-Type: application/json' \
  -d '{}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/rba_cash_rate. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Run it from npm

@pipeworx/mcp-rba is a local stdio bridge to the Pipeworx gateway: your MCP client starts it with npx, and it forwards every call to the same https://gateway.pipeworx.io/rba/mcp endpoint shown above. It needs network access; there is no offline mode.

{
  "mcpServers": {
    "rba": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-rba"]
    }
  }
}

Without a key it calls anonymously, capped at 50 calls a day. A free account raises that to 200 a day: get a key at https://pipeworx.io/signup and add "env": { "PIPEWORX_API_KEY": "<your key>" } to the server entry. The key is sent as a Bearer token, the same way a hosted connection sends it, so calls are metered and billed exactly as they are there.

It lists only this pack's tools by default. Set PIPEWORX_META_TOOLS=1 to also list ask_pipeworx and the other gateway-wide tools.

Behind an HTTPS proxy: Node's built-in fetch does not read HTTPS_PROXY by default. Newer Node releases honour it when NODE_USE_ENV_PROXY=1 is set; check your Node version's documentation.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Rba data" })

The gateway picks the right tool and fills the arguments automatically.

More

• Docs and guides
• pipeworx.io

License

MIT