Machine-Readable Access to Etla Research

llms.txt

llms.txt is a machine-readable index of Etla’s publications. It is placed at the domain root and lists economic forecasts, industry reviews, reports, briefs, and working papers — each with a Markdown link and a short summary. The file is intended for real-time reading by language models, so that the content is conveyed accurately and with correct source attribution.

Each publication is also available directly in Markdown by appending .md to the publication’s URL. Category and listing pages are available the same way (e.g. https://www.etla.fi/en/publications/reports.md).

MCP server

The Model Context Protocol (MCP) server gives AI agents direct, tool-based access to Etla’s content: filtered search, retrieval of any single item (of any type), and listing of publication series. The content is public.

• search-content
  Search all Etla content (publications, research projects, news, statements, columns, expert profiles, events). Filter by content type, publication series, tag, keyword, JEL code, author, language, or date; results are relevance-ranked.
• get-content
  Retrieve a single item’s full, type-specific metadata by id, slug, or URL — and optionally the Markdown body. Works for any content type; related items (research project, group, authors) are returned as linked references.
• list-series
  List all Etla publication series (e.g. Reports, Working Papers, Briefs, Suhdanne) — with each series’ publication count, latest date, and languages. A catalogue overview before a targeted search.

Gemini Enterprise

Gemini Enterprise requires OAuth authentication from MCP servers, which Etla’s direct MCP endpoint does not use. Etla therefore provides an OAuth bridge through which any Gemini Enterprise organization can connect Etla’s content to its own environment. (The direct endpoint /wp-json/mcp/etla is used as-is by e.g. Claude, Gemini CLI, and Antigravity.)

Connecting: in the Google Cloud console, Gemini Enterprise → Data stores → Create data store → Custom MCP Server (Preview) → enter the values above → Login → Create. Once the data store is Active: Actions → Reload custom actions → Enable actions, and connect the data store to your Gemini Enterprise app.

Requires a Gemini Enterprise license (Standard/Plus/Frontline). The organization’s administrator must also allow custom MCP data stores (organization policy) and grant Discovery Engine permissions. The content is public and read-only — the OAuth layer only satisfies Gemini’s mandatory authentication requirement, so the Client Secret is a shared, non-confidential value.

Machine-readable discovery

The MCP server’s metadata — endpoint, transport, capabilities, and tools — is published at the emerging .well-known paths, so that agents and aggregators can discover them automatically.

Use and attribution

Etla’s publications are public. When you use or cite the content — including in AI-generated responses — we ask you to mention the source, so that the reader can find the original publication and its authors. A recommended citation is provided with each publication.

Contact

Questions about machine-readable access, MCP integration, or AI use of content: