Kin
Install

Open-source developer alpha

Get Kin running in a few minutes.

Install the CLI, let Kin configure supported clients, initialize one repository, and ask the graph your first question.

Inspect release details

macOS and Linux terminal

curl -fsSL https://get.kinlab.dev/install | KIN_NO_SETUP=1 sh

Or with Homebrew

brew install firelock-ai/kin/kin

Review the installer first if your security policy requires it.

Agent setup prompt

Paste this into Claude, ChatGPT, Cursor, Gemini, or Antigravity and the agent installs Kin and registers its own MCP client.

Works with agent CLIs and IDEs that support MCP. macOS, Linux, and Windows, where native support is early and WSL2 is recommended.

Show the prompt
Please set up Kin for this project. Kin is the system of record for AI-written software.
It lets you ask the graph where code lives instead of searching files for it.

Work from the root of the repository we are setting up.

First install the CLI.

  curl -fsSL https://get.kinlab.dev/install | KIN_NO_SETUP=1 sh
  exec "$SHELL" -l

Then admit this repository into the graph. Kin reads the whole reachable Git history, so a
large repository can take a while. This cost is paid once.

  kin init

After init the daemon keeps preparing in the background, so the first agent calls on a large
repository can take noticeably longer to answer.

Now wire the agents. This detects the AI clients Kin supports and registers its MCP
server with each one it finds.

  kin setup --intent agent

Confirm it worked before going further.

  kin setup doctor

Check that the MCP clients line reports ok. Then restart your AI client completely, because
clients read their MCP configuration at startup and will not see Kin until they do.

After the restart you should see Kin tools available, including semantic_locate,
get_context_pack, and find_references. A good first request looks like this.

  Use Kin to explore this codebase. Run semantic_locate to find the main entry point, then
  get_context_pack on that file.

From here, prefer Kin's semantic tools over grep and raw file reads when you are exploring
this repository. Use semantic_locate to find symbols by meaning, get_context_pack to pull a
structured bundle around a file or symbol, and find_references to see what depends on an
entity. These read the graph directly and return what the repository actually recorded rather
than what a text match happened to hit.

One note on platforms. This works on macOS, on Linux, and on Windows. Native Windows
support is early and x64 only; use WSL2 on Arm.

First useful result

Four steps from install to context.

Run the graph on a repository you know. Kin reports its actual state before you depend on the answer.

  1. 01

    Install the Kin CLI

    Use the checksum-verified public installer for macOS or Linux. Windows has a separate PowerShell path.

    curl -fsSL https://get.kinlab.dev/install | KIN_NO_SETUP=1 sh
  2. 02

    Connect supported tools

    Run setup once. Kin reports which detected clients it configured and which still need attention.

    kin setup
  3. 03

    Initialize and verify

    Build the repository graph, then confirm its actual readiness before querying it.

    cd /path/to/repository
kin init
kin status
  4. 04

    Ask your first question

    Start with a task you know. Compare the answer with the repository and keep any reported gaps visible.

    kin locate "where are webhook retries handled"

Vector similarity is optional. Run kin embed when you want it, and budget for the model download and the local memory it needs.

Platform notes

Core support and filesystem projection are separate.

You do not need filesystem projection to use the CLI, daemon, or MCP tools.

macOS

Core supported. Filesystem projection may be limited by SIP or hardened executables.

Linux

Core supported. Filesystem projection requires a compatible glibc host.

Windows

Native Windows support is early and x64 only. Use WSL2 on Arm.

npm

npm i -g @kinlab/kin

Windows PowerShell

irm https://get.kinlab.dev/install.ps1 | iex

Your next step is a useful query.

After setup, ask Kin where a behavior lives or what depends on a change.

Open the documentation