This is a map of everything the Unstoppable desktop app can do, organized by what you're trying to get done. The app is home to two products that share one window: Unstoppable Code, for building software with AI agents, and Unstoppable Cowork, for everyday knowledge work in chat. Find the thing you want to do below and follow the link to the article that covers it.
New here? Start with the two overviews: What is Unstoppable Code? and What is Unstoppable Cowork?
Unstoppable Code (for developers)
Run AI coding agents — Claude Code, Codex, and more — in parallel, each in its own isolated worktree, with every step visible.
Get set up and oriented
Everything you need for your first hour: get in, learn the vocabulary, and run one task end to end.
- Getting access — sign in, get into early access, and install the desktop app.
- Core concepts & vocabulary — the handful of terms (workspace, worktree, plan, pipeline, hook) that make everything else click.
- Create your first worktree — your first task from prompt to result.
- The interface, at a glance — a quick tour of the pages, status indicators, and shortcuts worth learning early.
Work on many things at once
The habit Unstoppable Code is built around: start several tasks, let them run, and check each as it frees up instead of waiting on one.
- Working in parallel — run several worktrees at the same time and stay oriented while you do.
- Keeping your backlog flush — read the status indicators and queue work so your agents are never idle, especially before you step away.
Plan before you code
For anything meaningful, have an agent draft a plan you can review and refine before a line of code changes.
- Planning mode & plan tracking — generate a step-by-step plan, review it with diagrams and diffs, refine it, and track it as it changes.
- Executing a plan — turn a plan into work three ways (Pipeline, Implement, Copy), from fully automated to fully manual.
Automate the busywork
Chain agents, scripts, and quality gates into repeatable workflows, and let the app respond to events on its own.
- Pipelines: overview — what a pipeline is, how it runs, and where to build one.
- Anatomy of an implementation pipeline — the full lifecycle of a deep pipeline, from plan or task input to reviewed code.
- Reading a pipeline execution — follow a run, read its diagram and reports, and act on the result.
- Hooks: automate the busywork — fire an automatic response to events like a failed check, a merge conflict, or a new PR comment.
- Watchers — get pinged in a chat when a job you're watching finishes, succeeds, or fails.
- Automations & scheduling — kick off work on a schedule or at a set time.
Review and ship
Review code across more than one model before it ships. Dedicated how-to articles are on the way; for now, the review stages are covered inside Anatomy of an implementation pipeline and the What is Unstoppable Code? overview. Multi-model review across Claude and Codex, security and compliance checks, PR comments, and splitting a large change into separate pull requests all live in the app today.
Configure your setup
Bring your own tools and tune the app to your workflow: skills and plugins, MCP servers, environment-variable sets split by runtime, cross-agent compatibility between Claude and Codex, lifecycle and custom scripts, branch-naming and model settings, and dev-server detection with the in-app browser. These are available in the app now; step-by-step articles are being written.
Scale up and go remote
Push past a single machine: group several repositories into one workspace, run agents in the cloud, and monitor and approve work from any browser. Full audit trails, usage and cost statistics, a shared team feed, file viewing with handoff to VS Code, and privacy and cloud-sync controls round it out. These capabilities are live; their articles are on the way.
Unstoppable Cowork (for everyone else)
The everyday-work side of the platform — research, writing, planning, and getting things done in chat, organized into projects that keep their context. No coding or terminal required. Start with What is Unstoppable Cowork?, which covers the areas below; dedicated how-to articles are being written.
Start working
Set up a project — a container for related work — and start your first chat. Each project keeps its own notes, uploaded files, and the documents the assistant creates, so nothing needs re-explaining.
Connect your tools
Add plugins and connect the apps you already use (such as Slack or Linear) so the assistant can take action, not just give advice.
Manage your chats
Run several chats at once, fork one to branch in a new direction, archive the ones you've finished, and search across everything in your workspace.
Works across both products
- App Advisor — a natural-language operator for the whole app. Tell it what you want in plain language and it drives the interface — navigating, starting chats, creating worktrees, building pipelines, and setting up automations — across both Unstoppable Code and Unstoppable Cowork.
- Choosing models — pick the model and effort level that fit the task. The roster spans Claude, Codex, and more; an article on model strengths is on the way.