Claude Code from your phone: Remote Control and Happier, compared

What Remote Control does well, specifically

Every claim in this block is from Anthropic’s own Remote Control documentation, including the flag defaults. None of it is hedged, because a vague concession is not a concession.

Three ways into a running Claude Code session

Open the session URL in any browser, scan the QR code, or open claude.ai/code and find the session by name — Remote Control sessions show a computer icon with a green dot when they are online. With `claude remote-control` the QR is behind a spacebar press, and the VS Code command shows no QR at all; it posts the session URL into the conversation instead.

/rc adopts the Claude Code session you are already in

You do not have to decide in advance. Type /rc inside a running Claude Code session and it becomes remotely controllable, carrying over the conversation history — no restart, no losing the context you have built up over the last hour.

It reconnects after your laptop sleeps, and pushes when it needs a decision

If your computer sleeps or the network drops, Claude Code reconnects when it comes back online, and queues subagent and workflow status updates while the connection rebuilds so you get them once it recovers. Push notifications arrive when Claude decides one is warranted, or when it needs a decision from you to continue.

Server mode runs a pool of sessions from one process

`claude remote-control` stays running and serves sessions on demand. `--capacity <N>` caps how many run at once — the documented default is 32, which is a default rather than a ceiling. `--spawn` decides where they land: `same-dir` by default, so sessions share the working directory, or `worktree`, which gives each on-demand session its own git worktree while the session pre-created at startup stays put.

Included with Claude Pro, Max, Team and Enterprise

Remote Control is available on Pro, Max, Team and Enterprise. Two things the same page adds: it is a research preview, and on Team and Enterprise it is off until an Owner turns on the Remote Control toggle in Claude Code admin settings. API keys are not supported at all.

The five situations where Remote Control turns itself off

These are not limitations we found by testing. They are documented behaviour: in each case Remote Control is disabled or unavailable by design, and Anthropic says so.

1. 1. You authenticate with an API key.

   Anthropic’s docs state plainly that API keys are not supported for Remote Control, and that tokens created with setup-token or CLAUDE_CODE_OAUTH_TOKEN cannot establish a Remote Control connection.

   Happier treats API keys as a first-class auth path, not a downgrade. Store one as a connected service, bind it to a backend profile, and every device you own drives that session.

2. 2. Your traffic goes through a gateway or proxy.

   Remote Control is disabled whenever ANTHROPIC_BASE_URL points anywhere other than api.anthropic.com — which is exactly what every LLM gateway, proxy and internal routing layer does.

   Backend profiles exist to set ANTHROPIC_BASE_URL. Happier ships profiles for DeepSeek, Z.AI and MiniMax, and a custom profile can point Claude Code at your own gateway with your own variables.

3. 3. You run on Bedrock, Google Cloud’s Agent Platform, or Microsoft Foundry.

   Anthropic’s docs list Remote Control as not available on Amazon Bedrock, Google Cloud’s Agent Platform, or Microsoft Foundry.

   Be precise about this one: Happier ships NO first-party Bedrock or Microsoft Foundry support, and claiming otherwise would be the same kind of overreach this page is arguing against. What it ships is a custom backend profile that sets whatever environment variables your deployment needs — including the Bedrock and Foundry selectors — which you configure yourself and we do not test. Google Cloud is the exception: Gemini ships a Vertex AI profile out of the box.

4. 4. Your org turns telemetry off.

   Per Anthropic’s docs, DISABLE_TELEMETRY, DO_NOT_TRACK, CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC and DISABLE_GROWTHBOOK each disable the feature-flag evaluation Remote Control availability depends on.

   Happier has no equivalent kill switch, because it is not reading those variables to decide whether you are allowed to connect. Self-host the relay and the question does not arise.

5. 5. You are under Zero Data Retention or similar rules.

   Anthropic’s docs say organizations with compliance requirements such as Zero Data Retention can’t enable Remote Control, and that while it is connected the session transcript is stored on Anthropic servers.

   Happier accounts are end-to-end encrypted by default — the sync server holds ciphertext it cannot read. Run `happier relay host install` and the transcript never leaves hardware you control.

Remote Control drives Claude Code. Happier drives thirteen agents.

Will it run in your setup?

The eight facts that decide whether either thing works where you work. Two of them Anthropic wins outright. Those two are why the other six are worth reading.

What one client for 13 agents buys you

None of these are Remote Control doing its job badly. They are the things that only become possible once the client is not tied to one vendor’s agent.

Every agent in one inbox.

Claude Code’s remote opens in claude.ai/code or the Claude app. Codex reaches your phone through the ChatGPT app. Each vendor’s remote lands inside that vendor’s product, which is fine until you use two of them. Happier is one inbox instead: every session, on every agent, on every computer you have connected, in one list — with one place to answer the permission prompts they are all waiting on.

Mix and match agents inside one workspace.

A Codex session can hand the UI work to a delegate run on Claude while a Pi session digs through a stack trace, and you can steer any of them mid-run. Happier’s review, plan and delegate runs take a list of provider targets — not whichever provider you happen to be sitting in. A remote built by one vendor drives that vendor’s agent; Remote Control is a Claude Code feature, by its own documentation.

Choose an account, or pool several and let Happier fall back.

Connect more than one profile per service — a personal Codex subscription and a work one, a Claude setup-token, an Anthropic API key — and choose which one a session runs on at the moment you start it. Quota snapshots for Codex, Claude and Gemini profiles show as badges in the auth picker, so you can see what is left before you commit a long run to it. Nothing switches until you build a pool on purpose; once you have, it starts with automatic fallback on for the accounts a running session can actually move between, and there is a toggle on the pool if you would rather switch by hand.

Switch between the terminal and the app, mid-session.

Start a session in the app from your phone, then run `happier attach` on that computer and you are in that same session in a shell, with the history already there — for Claude Code and Codex inside tmux, for OpenCode through its own attach, and in all three you are in the agent’s real TUI. Or start it in the terminal and pick it up in the app. It reconnects to the session that exists rather than creating a second one, so nobody has to choose a side at the start of a task.

Start a session on any computer you own.

Sessions are created against a machine, so a Codex session can start on the VPS while you are on the laptop, and a Claude session on the laptop can message it and wait for it to finish. A live session can also move between machines — same session id, different computer — carrying its provider state and project directory with it, which works for Claude Code and OpenCode, and is experimental on Codex.

Mark the lines. Send them to any agent.

Open a file or a diff — the changed-files review, or the diff the agent just wrote into the transcript — and leave a comment on the exact line. Happier saves the path, the line, a snippet and a hash of that line’s contents, so the note still finds its code after the file moves underneath it. Pick which comments to send and they go in as structured review context: back into the same session, or into a new one on a different agent, because the comments are held against the workspace and any session you open there finds them waiting.

The other difference is where your conversation lives. Remote Control stores the session transcript on Anthropic servers while it is connected, per its documentation, and organisations under Zero Data Retention rules cannot enable it at all. Happier encrypts the transcript end to end by default, and happier relay host install puts the relay on hardware you own.

Each of the 13 agents has its own page — start here.

Trying it