Open Cranes

Claude Code

Add Cranes to Claude Code with the Cranes plugin, which brings in your thinking when you type @cranes and keeps your skills in each session, or with one command.

What you get

There are two ways to add Cranes to Claude Code:

With the plugin Connection only
Your thinking when you type @cranes Brought in before Claude starts on the message Claude calls Cranes when you ask
Your Cranes skills Copied into Claude Code at the start of each session Come in with your thinking when they fit
Reading and changing your thinking from a session Yes Yes
Saving When you ask, or say yes to an offer When you ask, or say yes to an offer

The plugin also covers Codex and Cursor; see Codex and Cursor.

Set up

With the plugin. It comes from Cranes’s public plugin repository, cranesintelligence/plugins.

  1. In Claude Code, run:

    /plugin marketplace add cranesintelligence/plugins
/plugin install cranes@cranes

  2. Restart Claude Code, or run /reload-plugins.

Connection only.

  1. In Cranes, choose Use in other AIs in the sidebar. The Claude Code card under Add an AI has this command with a copy button.

  2. In a terminal, run:

    claude mcp add --transport http cranes https://usecranes.com/connect/mcp

  3. In Claude Code, run /mcp and choose Cranes.

Sign in

Each sign-in opens Cranes in your browser. Check what it may use, choose whether to Keep my own ideas, and choose Allow.

With the plugin, there are two sign-ins, because the plugin’s automatic parts can’t use Claude Code’s own connection:

  1. Run /cranes:login. This one is for @cranes and your skills. Claude Code says in a line whether you’re signed in, and it renews itself. Until you sign in, the plugin adds nothing to your messages.
  2. The first time Claude uses Cranes for anything else, like saving or reading a page, Claude Code asks you to connect Cranes. You can also do it from /mcp.

Both show under Use in other AIs, the first as Cranes for coding agents. That first sign-in is shared on your computer with the plugins for Codex and Cursor, so you sign in once for all three.

Connection only: sign in when /mcp sends you to Cranes.

What’s automatic, and when you ask

With the plugin:

  • Your thinking when you type @cranes. Before Claude works on the message, the plugin asks Cranes what your thinking guides for it, and Claude shows what Cranes guided and where each part came from. @cranes on its own checks the work in front of Claude. When nothing applies, Claude says so in a line. Messages without @cranes never reach Cranes.
  • What it sends. Only for a message with @cranes: the message, plus a short description of the work (the repository and branch, whether you have uncommitted changes, and up to five of your earlier messages in the session that corrected Claude, word for word). Never your code, files, or Claude’s replies. If Cranes is slow or can’t be reached, your message goes ahead as usual.
  • Your skills. At the start of each session, your Cranes skills are copied into Claude Code’s skills folder as cranes- plus the skill’s name, so they load like any other skill. A copy you edit is never overwritten; Claude offers once to save your edit back to Cranes. Copies of skills you removed from Cranes are removed unless you edited them. Skills Cranes didn’t write are never touched.

Connection only: Claude Code calls Cranes when you ask: type @cranes, or ask what your thinking says.

Either way, saving waits for your yes. To skip typing @cranes in a repository, add the line under Have your AI use Cranes on its own on Use in other AIs to its CLAUDE.md. See When Cranes comes in and What connected AIs keep.

Claude Code Ask us