PelicanEvery Claude Code session on your computer, on your phone.
No wrapper, no relay, no account.
The sessions you started in the terminal, Claude Desktop or your editor, live, with their whole history. Continue any of them from your phone, approve Claude’s edits with the diff, or start new work.
npm install -g pelican-ccOpen source, MIT · No account · No telemetry · macOS and Linux
Every session, wherever you started it
- Claude Code writes it down. Every session it runs, in the terminal, Claude Desktop or your editor, goes into a transcript in
~/.claude. - Pelican reads it, live. One small process on your computer follows those transcripts, back past every compaction, and continues a session through the official Agent SDK, with your login, settings,
CLAUDE.md, skills and MCP servers. Claude Code talks to Anthropic, or your provider, as it always does. - Your phone opens Pelican. Over Tailscale or an SSH tunnel, signed in. Nothing sits in between.
A session another app is working on says which app, and when it waits for you there. Once it is idle, continue it here.
What you do at your desk, from your phone
Pelican runs Claude Code itself, so everything you set up for it applies. On top of that, what you would reach for in the terminal:
- Approve with the diff. Every edit and command waits for you, shown the way you would review it. Or pick a mode, as in Claude Code: Accept edits, Plan, Auto or Bypass.
- Steer or queue. Tell Claude something while it works, and it reads it at its next step; or line up what comes after this turn.
!commands. Run a command in the session’s folder. Claude reads the output with your next message.- Changes and files. Each changed file with its diff; tap a line to comment. Browse, read and mention any file.
- Visuals and pages. Charts Claude draws and pages it writes, in place, as still pictures until you tap Run.
- Rewind and branch. Put files back as they were before a prompt, or change an earlier prompt and try again on a branch.
- Every model, every command. Opus to Haiku, effort up to max, fast mode;
/commands and skills,@files and subagents, photos and files. - Notifications. When Claude finishes or needs you, and when a session waits for you in another app. Add Pelican to your Home Screen and it opens like an app.
Only you get in
Whoever can use Pelican can run commands as you, so it keeps everyone else out.
- Only this computer, unless you say otherwise
- It listens on 127.0.0.1, and every browser signs in, there too: with the link Pelican prints, a QR code that works once, or a password.
- Guarded against other websites
- Host and origin checks stop DNS rebinding and cross-site requests, and no other site may frame it.
- What Claude writes stays on the screen
- Pages and visuals render in a sandbox that cannot reach Pelican, as still pictures until you tap Run.
- Folders run nothing of their own
- The hooks, MCP servers and git filters a folder brings wait until you trust it.
Other ways to reach Claude Code
As each project documents it, checked on 5 November 2026.
| Tool | Your sessions | Between phone and computer |
|---|---|---|
| Pelican | All of them, from any app, with their whole history | Nothing: Tailscale, an SSH tunnel or your proxy |
| Remote Control | Those with Remote Control turned on; Claude plans only | Anthropic’s servers, which keep the transcript while connected |
| Yep Anywhere | Resumes those from the CLI, VS Code and desktop apps | Nothing, or its end-to-end encrypted relay |
| Happy, HAPI | Those started in Happy or HAPI | An end-to-end encrypted relay, or with HAPI your own tunnel |
| CloudCLI | Every session in ~/.claude | Nothing on your network, or CloudCLI Cloud |
| SSH and tmux | The ones running in tmux | Nothing |
Remote Control can answer a prompt waiting in your terminal; Pelican cannot. More in the docs
Before you install it
Does it work with my Claude plan?
Pelican runs Anthropic’s own Claude Code, unmodified, on your computer, signed in the way you signed in to claude; it never sees your credentials, and it is for your own use. Signed in with a plan, what you send from Pelican draws from that plan’s limits. The details
Why not Remote Control?
Use it if you are on a Claude plan and want to answer, from your phone, the prompt waiting in your terminal: Pelican cannot. Pelican shows every session without turning anything on, keeps nothing on anyone’s servers, and works with any login, API keys and cloud providers included.
Why not SSH and tmux?
That works, and needs nothing new on the computer. Pelican is for what a terminal on a phone does badly: long transcripts, diffs, approvals in one tap, notifications, charts and pages, and history across compactions.
What if my computer sleeps?
A computer that is asleep runs nothing, and Pelican cannot wake it. While Claude works, Pelican keeps it from going to sleep on its own; keep it awake yourself to reach Pelican at any time.
Does it cost anything, or phone home?
No. Pelican is free and MIT licensed, with no account and no telemetry. It talks to your browser and, for notifications, to the browser’s push service, with the content encrypted.
Why is the install about 150 MB?
The Agent SDK, which Pelican runs Claude Code through, brings its own copy of Claude Code. The live demo needs no install at all.
One command, then your phone
-
Install it
On the computer where you use Claude Code, with Node 22.12 or newer.
pelicanprints the link that signs this browser in.npm install -g pelican-cc pelican -
Open it on your phone
With Tailscale on the computer and the phone. Scan the QR code, and Pelican opens there, signed in.
tailscale serve --bg 7860 pelican phone -
Keep it running
It starts when you log in, and again if it stops.
pelican service install
About 150 MB, most of it the copy of Claude Code the Agent SDK brings. Other ways in, or try the live demo first.