No. CodeAgent Mobile isn't built for writing code on a small screen. It's a supervision layer — your agent does the work in your IDE or a Codespace, and your phone is how you approve, redirect, and stay in the loop without being at your desk.

Claude Code, GitHub Copilot, Cursor, Windsurf (Cascade), JetBrains AI Assistant, Junie, OpenAI Codex, Gemini CLI, Aider, and CodeRabbit — whichever agent your IDE or terminal already runs.

No. CodeAgent Mobile pairs with the agent you're already using. Install the VS Code or JetBrains plugin (or codeam-cli for terminal use), scan a QR code, and your phone becomes a supervision surface for that existing session.

Yes. Pick a Jira, Linear, or Azure DevOps ticket and the session assembles itself: CodeAgent detects the repo the ticket maps to, wires in the tools the ticket links, and packs the context — title, description, and acceptance criteria — into an agent-ready session. Choose an intent (Implement, Test, Investigate, or Document) to frame the work, and when it lands the agent comments the PR back on the ticket, so nothing falls out of your tracker.

Yes. Pick a Slack thread and CodeAgent summarizes it, then creates a ticket in your tracker for you — assigned, story-pointed, with a description and acceptance criteria drawn from the conversation — detects the repo it belongs to, packs the context, and launches an agent-ready session. When the work lands it closes the loop on both sides: it comments the PR on the ticket it created and replies back in the original Slack thread. It's the one-step-earlier sibling of starting from a work item — that flow needs a ticket to already exist; this one files it for you.

Yes. The PR Command Center shows the GitHub PRs you're asked to review and the ones you opened. Before you act, you can read each changed file as a full diff in the app's built-in editor — the same viewer the IDE uses — so you're actually reviewing the change, not rubber-stamping it. As a reviewer you can approve, request changes, or comment; as the author you can merge or close, and if a PR you authored has merge conflicts you can hand them to an agent to resolve and update the PR — all posted straight to GitHub. You can also send one of your linked coding agents to review a PR: it checks out the branch, leaves inline comments where it finds issues, and approves or requests changes, then you get a push notification with the verdict. It's the review-and-ship end of the same loop that starting from a ticket or a Slack thread opens.

If the agent is running locally, the session pauses with your machine. If you're running through a Codespace (via codeam deploy or the app), the agent keeps working in the cloud — you can close your laptop entirely and get notified when it's ready for review.

You choose. CodeAgent Mobile runs your agent on any of three execution planes, all driven from the phone: (1) Local — pair the IDE or terminal on your own machine; (2) Cloud — launch a GitHub Codespace from the app so the agent runs in the cloud and keeps working with your laptop closed; or (3) Self-hosted — deploy the agent to your own server, VM, or Docker box (Pro) and supervise it from mobile. It's not just an app to pair a local IDE — it can provision and run the agent for you in the cloud or on your own infrastructure.

Yes. The Self-hosted plane (Pro) deploys your agent to a server, VM, or Docker box you own: CodeAgent Mobile installs codeam-cli on the host, authenticates your agent, and supervises it from the app with live host telemetry — bring-your-own-infrastructure, no dependency on GitHub Codespaces.

Yes. Tap Preview on a paired session and CodeAgent Mobile detects how to start your project's dev server, brings it up, and serves it over a secure public tunnel — so the running app opens right inside the app (or as a QR for Expo). No manual port-forwarding, and your machine never has to be reachable from the public internet. You can also view and edit the project's environment variables from the app and restart the preview to pick them up.

Yes. From a session's add-ons you can turn on a cost-saving proxy that compresses the tokens your agent sends to the model provider on long autonomous runs, and the app reports the savings back to you. It's optional and off by default.

Yes. Team Spaces let you share a live agent session with up to 5 collaborators (Pro). Everyone sees the same session, can redirect it, and the agent attributes prompts by name — no screen-share required. A shared Beads panel tracks what the agent is working on as issues (created, in progress, closed) so the whole team stays in sync on the same task graph, and team memories let an owner share conventions that carry into every member’s agent automatically.

Yes. The Free plan includes 5 agent tasks per day, 2 paired devices, 1 linked agent, and access to every agent your IDE already runs. No credit card required.

Pro removes all daily caps — unlimited agent tasks, AI Insights, Smart Composer, and @codeagent mentions — plus unlimited, concurrent cloud workspaces (Codespaces + CodeAgent Boxes, vs 1 active on FREE). It also adds: up to 5 paired devices; Team Spaces for up to 5 collaborators (with the shared Beads task graph); the self-hosted deploy plane (run the agent on your own server, VM, or Docker box); Agent Toolkit integrations like Atlassian (Jira + Confluence); the built-in cost-saving Headroom compression that cuts token waste on long runs; persistent cloud previews; a full, exportable agent-action audit log (every prompt, approval, file change, and deploy); priority infrastructure that stays responsive under load; priority support (replies under 24h on weekdays); and first access to new agents and integrations. $19.99/month or $191.99/year.

Yes. codeam-cli runs your agent in the terminal over the open ACP protocol — Claude Code, OpenAI Codex, Gemini, and other ACP-compatible agents — so it's not tied to one vendor. If you can run Node.js, you can run codeam, including over SSH or inside containers.

You can — and if you enjoy running a server, registering a domain, configuring a tunnel and keeping port-forwarding secure, it works well. CodeAgent Mobile is for everyone who'd rather skip all of that: the same remote supervision of your existing agents, but with nothing to host, expose, or maintain. Pair once and you're driving from your phone in under a minute — no infrastructure, and your machine never has to be reachable from the public internet.

Your code stays where your agent runs — on your own machine or your own GitHub Codespace; we never clone your repository to our servers. CodeAgent Mobile acts as a relay: the prompts you send and the agent's streamed output (including diffs) pass through our backend to reach your phone, buffered only briefly and not retained. Team Space chat messages are stored so the conversation persists for your team. Your agent credentials are encrypted at rest.

Yes. Tap the Voice button in the composer — it opens a Compose | Talk switch. Compose turns what you say (or an attached screenshot or file) into a clear, structured prompt in your own language before it's sent; Talk (live, PRO) lets you supervise your coding agent by voice, hands-free — ask it to review, approve, or merge a pull request, catch you up on what your agent just did or is working on, send it a new task, or spin up / wake / stop a Codespace, and it works and narrates progress back to you.

Yes. Pro is month-to-month or annual, cancel anytime, with a 7-day money-back guarantee on your first purchase.