A reader’s map of the great novels — never a step ahead of you.

Great novels reward attention, but they also ask a lot of memory — who was that minor character forty pages ago, how does this place connect to that one, which thread is about to pay off. The usual reference works answer those questions by spoiling the ending. LitGraph exists to give readers the same orientation a careful re-reader has, without ever revealing what is still to come.

We start with the texts the world already owns. Every book on LitGraph is in the public domain, produced from carefully proofed Standard Ebooks editions and released under CC0. The prose was never the secret — the structure was, and that is what we map.

Each reader has a position in the book — a “ceiling” that marks the furthest point they have read. Everything LitGraph shows you is clipped to that ceiling. A character’s connections, the places they have been, the threads they appear in, even the inline cross-references in the text are revealed only up to where you actually are. Read ahead in the prose if you like; the graph waits for you.

The public pages you may have arrived from work the same way. Every answer on a character or book page is bounded to an early reading position and labelled as such, so a search engine, an AI assistant, or a curious browser can cite a fact about a novel without anyone stumbling onto its ending.

Building a spoiler-safe map by hand would take an editor weeks per book, so we do it in stages. First, an ingest pass reads each chapter in order and extracts the characters, places, and relationships it observes — strictly chapter by chapter, so nothing from later in the book contaminates an earlier position.

A reconcile pass then resolves the ambiguities a single chapter cannot: the same person under a title, a nickname, and a formal name is merged into one entity; coincidental name collisions are kept apart; weak or unsupported links are dropped. The result is a clean, position-aware graph rather than a pile of mentions.

Finally, an editorial review closes the loop. A human checks the extracted graph and writes the public-facing prose — the short citable answer, the longer biography, the frequently-asked questions — confirming that every published sentence respects the spoiler boundary before it goes live. The byline and reviewer on each page name the people accountable for that copy.

Experience. LitGraph is built by readers who have actually worked through these novels page by page, not by scraping summaries. The position-bound graph is the same tool we use to read.

Expertise. Entities are extracted, reconciled, and cross-checked against the canonical text, then linked to authoritative references — Wikidata, Wikipedia, Standard Ebooks, Project Gutenberg — so a claim can always be traced back to its source.

Authoritativeness.Every public page declares what it is bounded to, who authored it, who reviewed it, and when it was last updated. The same attribution appears both in the visible “About this page” block and in the page’s structured data, so the two never disagree.

Trust. We publish from public-domain sources, we never invent facts a book has not established, and we hold every sentence to the spoiler boundary. If a page cannot make a claim without spoiling, it does not make the claim.

Browse the catalog to find a novel, then open a character guide to get oriented before or during your read. When you are ready to read the book itself, sign in and the in-app graph will track your position automatically, unlocking each new character and connection exactly when the book introduces them.

Reading is free for every public-domain title. You stay in control of the pace, and the map never gets ahead of you.