Skip to content
MCP serverOverviewDocs menu

MCP server

6 read-only tools and the library as resources, over Streamable HTTP with no sign-in.

POST/mcp

The server speaks the Model Context Protocol, so an assistant can search the library, read texts and use their analyses as tools. It is built on the official TypeScript SDK and reads the same data as the REST API.

Connection

  • Transport: Streamable HTTP at https://interlinea.aeterna-institute.org/mcp. POST carries every message; GET and DELETE answer 405.
  • No authentication and no session. Each request stands on its own, so any number of clients can connect.
  • Protocol: the 2026-07-28 revision, and 2025-era clients through stateless serving. Today's Claude, ChatGPT and IDE clients connect as they are.
  • CORS is open, so a client in the browser can connect too.

Quickstart has the setup for each client.

Instructions

The server sends these instructions when a client connects. Most clients pass them to the model.

instructions
Interlinea is a free library of Latin and Ancient Greek texts with word-by-word analysis (https://interlinea.aeterna-institute.org).
Start with search_library or list_passages, read with get_text (follow next.url to keep reading), and use get_analysis, lookup_word and get_context for grammar, vocabulary and background.
Every result carries a citation, a canonical url, a CTS URN where the edition has one, and the edition's credit and licence. When you quote or reuse a text, give the citation and url and credit the edition as its credit.text says. Each text keeps its own licence (credit.license): CC BY-SA, CC BY, or, for The Latin Library's texts, used with that site's permission and not openly licensed. Interlinea's analysis notes are CC BY-SA 4.0; they do not change the licence of the text they analyse.
Analyses and notes are stored ones only: nothing here generates new ones, so some sections return words: null or section: null. Open the url in Interlinea to have them made.
The same data is available as JSON at https://interlinea.aeterna-institute.org/api/v1 (OpenAPI: https://interlinea.aeterna-institute.org/api/v1/openapi.json). Docs: https://interlinea.aeterna-institute.org/docs.
Rules: Say who you are: send a User-Agent that names your agent or product and a way to reach you. Do not pose as a browser. Follow robots.txt. Use the MCP server, the JSON API, the Markdown pages and the sitemaps. Do not crawl /library/search or other filtered HTML pages: /api/v1/search answers the same questions. Keep to 120 requests a minute in total, however many addresses you send from. Do not spread requests across addresses or proxies to get past a limit. On 429 or 503, stop and wait at least a minute before you try again. Reuse what you have already fetched; the answers change rarely. Keep to each text's licence and credit, and to the law that applies to you and to us: copyright, data protection and computer misuse law among them. Traffic that breaks these rules is blocked without notice. If you need more than this allows, write to us first.

Tools

ToolWhat it does
search_librarySearch Interlinea's library of Latin and Ancient Greek texts by author, work title or topic (war, love, the gods, friendship, death, rhetoric, letters, comedy).
get_textRead a section of a Latin or Greek work, or a graded passage, in the original.
get_analysisThe stored word-by-word analysis of a text: for every word its lemma, dictionary entry, part of speech, Universal Dependencies features, head and relation, and meaning in that sentence, plus an English translation of each sentence.
lookup_wordLook up a Latin or Greek headword (amo, arma, λόγος) in the stored analyses: its dictionary entries, the meanings it had in context, and example sentences with translations, citations, urls and licences.
get_contextStored notes on where a text sits in its work: what happened before, what happens here (never a translation), the people, gods and places named, what comes next, and a few lines about the work.
list_passagesThe graded path: short Latin and Greek passages chosen for learners, from level 1 (Genesis, Aesop) to level 5 (Tacitus, Thucydides, Sappho), each analysed word by word ahead of time.

Every tool is read only and says so: readOnlyHint is true, destructiveHint is false, idempotentHint is true, openWorldHint is false. A client can call them without asking for confirmation.

Results

A tool answers twice over, as the protocol asks. structuredContent holds the result as JSON that matches the tool's output schema, and content holds the same JSON as one text item for clients that only read text.

Errors

A reference that isn't in the library, or isn't a text, comes back as a normal result with isError: true and a message that says what to pass instead:

Arguments
{
  "ref": "/library/latin/virgil"
}
Response
{
  "result": {
    "content": [
      {
        "type": "text",
        "text": "Virgil is an author, not a text. Pass one of their works, e.g. https://interlinea.aeterna-institute.org/library/latin/virgil/aeneid, https:/…"
      }
    ],
    "isError": true
  },
  "jsonrpc": "2.0",
  "id": 1
}

Resources

The graded passages and the best-known works are also listed as resources, and any work or section can be read as one. See Resources.