Be the first to get access to new connectors. Get early access →
ConnectorBay Early access

Stacks — connector documentation

  • MCP endpoint: https://stacks.connectorbay.com/mcp
  • REST API: POST https://stacks.connectorbay.com/v1/tools/{tool_name}, described by the OpenAPI 3.1 spec at https://stacks.connectorbay.com/openapi.json
  • Transport: Streamable HTTP, stateless, MCP specification 2025-06-18 or later
  • Authentication: none required. An optional Authorization: Bearer <key> raises rate limits.
  • Health: https://stacks.connectorbay.com/health · Tool schemas: https://stacks.connectorbay.com/tools.json
  • Publisher: Unify Dynamics LLC · connectorbay.com/stacks · agent@connectorbay.com

What Stacks is for

Questions about books, authors and archives, answered from open catalogues instead of memory: who wrote something and when, what a book is about and how readers rate it, which edition an ISBN is, an author's life and works, where to start on a subject, whether a free legal ebook or audiobook exists, the actual text of a public-domain book, what the Internet Archive holds on a topic, and what a web page looked like years ago.

What it does

Open Library: search 40 million records by words, title, author, subject, language or year range, with ratings, edition counts, covers, ISBNs and ebook availability; a work in full (description, subjects, places, people, rating distribution, editions); an edition by ISBN (publisher, date, pages, identifiers); an author with dates, bio, links and top works; a subject page (top works, authors, related subjects); trending titles. Project Gutenberg: the 78,000- title catalogue bundled and searched locally by words, topic, language or the year an author was alive, with download links in every format; and the text itself — chapter list with offsets, any passage, or every place a phrase occurs. LibriVox: free public-domain audiobooks with runtime and links. Internet Archive: search texts, audio, film, software and images by words, collection, creator and year; an item's metadata and files; the Wayback Machine's closest snapshot of a URL.

What it is not

Stacks sells nothing and holds no copyrighted full text: passages come only from Project Gutenberg's public-domain editions (public domain in the United States; check local law elsewhere). Open Library's data is community-maintained and a title can have duplicate records; Wayback captures are whatever the Archive crawled.

Tools

Open Library

search_books — Search Open Library's catalogue (40M+ records): title, authors, first publication year, edition count, ISBN, cover, subjects, reader rating, median pages, and ebook availability. Use for "who wrote…", "books about…", "novels by Ursula Le Guin sorted by rating", "free ebooks on gardening". Args: query, title, author, subject, language, year_from, year_to, sort, ebooks_only, max_results.

book — One book (work) in full: description, authors, subjects/places/people/times, reader rating with distribution, cover, links — plus optional editions. Accepts a work key, an ISBN, or a title (+author). Use for "tell me about Dune", "what is The Left Hand of Darkness about", "how is Sapiens rated". Args: key, title, author, include_editions.

isbn_lookup — An edition by ISBN: title, publisher, publish date and place, pages, format, languages, series, ISBN-10/13, LCCN/OCLC/Goodreads ids, cover, and the parent work key. Use for "what book is ISBN 978…", "who published this edition". Args: isbn.

author — An author: life dates, bio, alternate names, photo, external links (Wikipedia etc.) and their most-published works with ratings. Use for "who is Frank Herbert", "what did Chinua Achebe write", "when was Le Guin born". Args: name, key.

subject — Explore a subject: most-held works, top authors, related subjects, people and places. Use for "classic science fiction to start with", "who are the main authors on stoicism", "subjects related to permaculture". Args: name, ebooks_only, max_results.

trending_books — What Open Library readers are logging and reading right now — trending works for the chosen period. Use for "what books are trending", "popular books this month". Args: period, max_results.

Public domain

gutenberg_search — Search Project Gutenberg's public-domain catalogue (78,000 free ebooks, bundled and searched locally): title, authors with life dates, language, subjects, bookshelves, popularity (30-day downloads where known) and epub/kindle/html/text/cover links. Use for "free ebook of Frankenstein", "popular Gutenberg books in French", "children's books on Gutenberg", "detective novels by authors alive in 1900". Args: query, topic, language, author_alive_year, max_results, page.

gutenberg_book — One Gutenberg ebook: catalogue metadata and download links in every format, plus a short summary and live download count when Gutendex answers in time. Use after gutenberg_search, or with a known number ("Gutenberg #1342"). Args: book_id.

book_text — Read a public-domain book's actual text from Project Gutenberg: the chapter list with offsets, a passage from any offset, or every place a phrase occurs (with context). The Gutenberg header/footer is stripped. Use for "the first page of Frankenstein", "chapter list of Moby Dick", "where does 'Reader, I married him' appear in Jane Eyre", "read Pride and Prejudice chapter 3". Args: book_id, find, offset, length, chapters_only.

audiobooks — Free public-domain audiobooks from LibriVox: title, authors, language, total runtime, number of sections, description, and links (LibriVox page, Internet Archive page, RSS feed, zip of MP3s, source text). Use for "is there a free audiobook of Dracula", "audiobooks by Jane Austen". Args: title, author, max_results.

Internet Archive

archive_search — Search the Internet Archive (books, audio, film, software, images, web): identifier, title, creator, year, mediatype, collections, downloads, description, subjects, size/runtime and link. Use for "old public-domain films about atomic safety", "1920s jazz recordings", "scanned copies of Dracula", "old-time radio shows". Args: query, mediatype, collection, creator, year_from, year_to, sort, max_results, page.

archive_item — One Internet Archive item: full metadata (title, creator, date, publisher, subjects, language, licence, uploader, description), file formats present, and the largest original files with direct download links. Use for "what's in this archive item", "download link for the PDF of…". Args: identifier.

wayback — The Wayback Machine's closest archived snapshot of a URL (optionally near a date): snapshot link, capture time, HTTP status, and the link to all captures. Use for "what did example.com look like in 2010", "is there an archived copy of this page". Args: url, date.

Conventions

  • Read-only. Every tool reads public data or computes; nothing is written anywhere and nothing about the user is stored.
  • Result envelope. ok: true with data, or ok: false with a structured error (bad_input, not_found, upstream_error). degraded lists optional sources that were unreachable.
  • Rate limits. Free tier 60 tool calls/minute and 300/day per user (keyed on the user identifier supplied by the host agent, else client IP); Pro keys raise this. Over-limit calls return HTTP 429 with Retry-After. initialize and tools/list are never counted.

Support

agent@connectorbay.com · connectorbay.com/stacks/support