mcpField
K
Koragraph

Koragraph

Local
koragraph.inDeveloper Tools15Updated 6w ago

About Koragraph

Koragraph builds a local code graph and code-anchored memory across multiple repositories for coding agents. Its tools are available through MCP or the CLI, with a Claude Code hook to add relevant graph context each turn.

Key features

  1. 01Cross-repository code graph
  2. 02Code-anchored persistent memory
  3. 03Blast-radius and call-graph queries
  4. 04Nine tools via MCP and CLI
  5. 05Context-sized query output

Use cases

  1. 01Tracing dependencies before editing
  2. 02Exploring code across repositories
  3. 03Reusing fixes and project rules
  4. 04Adding graph context to Claude Code sessions

MCP support

Koragraph exposes nine code-graph and memory tools to MCP clients. The same tools are also available from its CLI.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"in-koragraph-koragraph": {
"command": "npx",
"args": [
"-y",
"koragraphmcp"
]
}
}
}
Run commandnpx -y koragraphmcp

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npmkoragraphmcp0.1.4

Tools

9

Saved tool list · updated 2w ago

ParameterTypeRequiredDescription
querystringYesA symbol name or a short plain-English description of what you want to understand or change.
detailenumNoconcise (default): a one-line answer plus the ranked head, file:line kept on every row. full: every row and every field.concise | full
project_idinteger|stringNoOptional scope: a repository or project name exactly as overview prints it, or its numeric id. Omit when the graph holds one project.

Details

From the project README
koragraph

The ultimate context layer for your AI coding agent.

Completely free. Runs entirely on your machine.

A local knowledge graph of every repo you point it at, plus a memory layer that knows when the code it remembers has changed.

npm License: BUSL-1.1 Node.js >=22 MCP compatible Website

Install

npm install -g koragraphmcp
koragraph ingest /path/to/repo        # point it at as many repos as you like
claude mcp add koragraph -s user -- koragraph mcp

koragraph doctor checks the whole chain and prints the exact line for any editor.

Two ways to use it from here: your agent calls the tools over MCP, or you (or a script, or a Claude Code hook) run them straight from the shell, for example koragraph explore "how does checkout work" or koragraph blast src/pay.js. Same graph either way.

Then run KORAINIT in your editor

This is the step that lights up the memory layer, so don't skip it. In your coding agent (Claude Code, Cursor, Windsurf, whatever you use), tell it:

"Read KORAINIT.md and follow it."

It imports your existing CLAUDE.md / AGENTS.md into koragraph's memory, anchored to the code each rule is about, flags any rule that points at code that's already gone, and finishes wiring koragraph into the session. Run it once per project.

Prefer zero setup? Hand the whole thing to your agent instead. Paste this from the repo you want indexed, and it installs, indexes, wires itself in, and runs KORAINIT for you:

Install koragraph and set it up for this repo, then report back:
1. npm install -g koragraphmcp
2. koragraph ingest .
3. koragraph doctor, then run the `claude mcp add …` line it prints (or wire `koragraph mcp` into
   my editor's MCP config).
4. Find KORAINIT.md in the installed package (`npm root -g`, then koragraphmcp/KORAINIT.md) and
   follow it: it imports my existing CLAUDE.md / AGENTS.md into memory, verified against the code.
Report: nodes/edges indexed, whether the MCP connected, and KORAINIT's summary.

Works with

Claude Code    Cursor    GitHub Copilot    Windsurf    Zed    Cline    Continue    OpenCode    Codex    Antigravity


Adoption is highest on Claude Code: its hooks let koragraph put the graph in front of the agent on every turn, so it actually gets used instead of falling back to grep. Other editors reach koragraph over MCP, where the agent decides when to call it.


One map of everything you've built

Koragraph turns every repo you point it at into one live graph: a resolved structure, with an edge only where there really is one.

  • Every service, file, endpoint, and table.
  • The real calls between them, resolved against the syntax tree, not guessed by name.
  • The edges that cross a repo boundary, where single-repo tools go quiet.
koragraph serve
koragraph serve: code (cool nodes) and its memory (gold diamonds), one local page.

A memory that knows when the code changed

Everything your agent learns is saved locally, anchored to the exact declaration, not a line number. Rename the code and the fact follows; delete it and the fact retires. No CLAUDE.md to maintain, and it survives every upgrade.

  • Failures and their fixes. The fix surfaces before you hit the same wall twice.
  • Hazards, rules, corrections. Stated once, honoured every session after.
  • The "let's push this to Wednesday." Deferrals and half-finished intent, parked against the code and surfaced the moment you're back.
Memory anchored to the code it is about

Measurably the most accurate, scored by the compilers

Every graph is graded by an independent compiler front-end (CPython ast, go/parser, Roslyn, syn, tsc, Ripper, php-ast, ctags, swiftc), never the tool under test. koragraph wins 34 of 39 language×plane cells, and leads precision on all 39.

Call-graph recall, the hardest plane:

LanguagekoragraphCodeGraphGitNexusGraphify
JavaScript87.639.441.945.1
TypeScript70.655.250.244.2
Python86.185.154.653.3
Go89.185.063.865.3
Java86.073.168.653.5
C#83.679.850.554.4
PHP88.988.077.131.7
Ruby51.541.328.740.8
Rust68.776.548.947.6

Reproduces from test/benchmark: pinned corpus, pinned competitor versions, pinned oracles. Full breakdown in test/benchmark/REPORT.md.


The tools

Nine tools, reachable two ways: as MCP tools in your editor, and as CLI commands in your shell. On Claude Code a hook also calls the CLI to push the relevant slice into context on every turn, so the graph gets used whether or not the agent thinks to ask. Every tool returns file:line, never pasted source. Exactly one writes.

ToolShellWhat it's for
blast_radiuskoragraph blast <files…>Run before editing. Everything that depends on this, across repos, and what has no test coverage.
explorekoragraph explore "<symbol or phrase>"Ranked declarations, source, callers, callees, in one call.
search_codekoragraph search <name>Find a name in the graph, not raw text.
neighbourskoragraph neighbours <symbol>Callers and callees of one symbol.
changes_withkoragraph cochange <symbol>What has historically shipped together with a symbol.
file_symbolskoragraph symbols <file>Declarations in one file.
overviewkoragraph overviewOrient on turn one, across every repo.
recallkoragraph recall <symbol>A past failure and its fix, a hazard, an open loop.
rememberkoragraph remember "<fact>"Save a durable, code-anchored fact. The only writer.

Every query verb takes --format compact|json|paths and --budget <tokens>, so a result can be trimmed before it ever enters an agent's context. Plus the operational CLI: ingest, serve, status, doctor, diff, trace, hooks, practice, report.


Languages

c · c++ · c# · go · java · javascript · php · python · ruby · rust · swift · typescript / tsx, fully resolved, tree-sitter throughout. Plus experimental kotlin · scala · elixir · solidity · vue · zig · objective-c · ocaml · rescript.


License

BUSL-1.1. Free to use, including at work. Converts to Apache-2.0 on the change date.


Built by Akhil Katakam · Ethan Faleiro  ·  [email protected]

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.