For AI agentsA Markdown index of these docs is at /llms.txt. Every page is also Markdown: add .md to its address.

Docs

LLMGRAPH docs

LLMGRAPH answers your coding agent's questions about your code. Every answer says what is certain, what is open, and what it can't see. Start from what you want to do.

Set up with your agentAdd the server, sign in once, and check that your agent sees the five questions.Start here →Read an answerWhat the five labels mean, and how to read each line of an answer.Read →What it can't seeLanguages, plan limits, and the blind spots an answer names.Read →Account, licence and privacyThe trial, devices, working offline, and what leaves your machine.Read →

Quickstart

  1. 1
    Add the server to your agent

    For Claude Code, one command. Other agents.

    $claude mcp add --scope project llmgraph -- npx -y llmgraph-mcp
  2. 2
    Sign in once

    Run this, or just ask your agent something: its first call gives you a sign-in link and a code.

    $npx -y llmgraph-mcp login
  3. 3
    Ask your first question

    Ask your agent who calls a function before it edits it. The answer starts with llmgraph · .

The five questions

Your agent picks one of these questions. The descriptions are the engine's own words.

who_callsEvery place that calls this function, plus calls we can't see.what_breaks_if_changedEverything that depends on this, directly or indirectly, and what we can't see.check_my_changesAfter editing: callers you didn't update, and anything we can't see.what_is_unknownWhere our view is blind (dynamic calls, unknown types, unreadable files) and why.explainWhy we said this: the evidence behind one answer line.

Docs for agents

/llms.txtAn index of every page, one line each.
/llms-full.txtEvery page in one file.
/docs/…/page.mdAny page as Markdown: add .md to its address.
For agentsthis page as Markdown/docs.mdllms.txt