“Which of these nine terminal tabs is beeping?” Solo is a desktop app born from exactly that daily frustration. Released in February 2026 by Try Hard Studios, LLC in the US, it gathers AI coding agents such as Claude Code and Codex together with long-running processes like dev servers and build watchers into a single, visible workspace. It is a native app built with Tauri, so it is lighter than Electron-based tools. The official site states plainly that it is “Not an IDE. On purpose.” — rather than replacing your editor or your agents, it sits alongside them and keeps track of what is running.
Key Features
- Agents and dev processes on one screen: Agents, dev servers, builds, and shell sessions all appear in the same dashboard, so you can tell at a glance which is running, which is waiting for input, and which has crashed. No more cycling through tabs to find out.
- Automatic restart of dead processes: When a command you have marked as trusted exits unexpectedly, Solo relaunches it automatically. That prevents the classic case of a dev server quietly dying without anyone noticing.
- Exposes state to agents as an MCP server: Solo itself acts as an MCP (Model Context Protocol) server, handing process state, logs, and project information to agents over a local stdio connection. Agents can also start, stop, and restart processes, or wait for a port to bind.
- No vendor lock-in: You are not tied to any one AI. CLI agents such as Claude Code, Codex, Gemini CLI, Copilot CLI, and Kimi CLI connect with one click, and any other command can be registered as well.
- Coordination across multiple agents: Shared todos, Markdown scratchpads, locks, and timers support workflows where one lead agent directs several worker agents inside the same project.
- Share configuration through your repository: Commit
solo.ymlto the repo and anyone on the team who clones it starts up with the same process setup. - A properly built native terminal: GPU-accelerated rendering, ANSI support, search, and clickable links.
Pricing
| Plan | Price | What you get |
|---|---|---|
| Free | $0 (forever) | Up to 4 projects and up to 20 processes total. All features, including MCP integration |
| Pro | $99/year | Unlimited projects and processes, activation on 3 devices, priority email support, 30-day money-back guarantee |
| Team | From $99/year per seat | Per-seat price decreases with volume (down to $69/year per seat at 51+ seats). Equivalent to Pro |
Pricing is current as of August 2026. Check the official site for the latest details.
Pros & Cons
✅ Pros
- You can see the state of every agent and process on one screen, so no time is lost hunting for whichever one is beeping
- No lock-in to a particular AI vendor; the CLI agents you already use keep working as they are
- Built with Tauri and lightweight — it does not feel like launching a second IDE
- Because agents can read process state and logs over MCP, you spend less time telling them “the server is down”
- The free plan has no feature restrictions, so it is genuinely usable within its limits (4 projects, 20 processes)
⚠️ Cons
- Current public builds cover macOS and Windows only; Linux is still listed as coming soon
- Paid plans are annual only — there is no month-by-month option for a short trial
- It is neither an editor nor an AI, so it does not complete your development setup on its own; it is a layer added on top of your existing tools
- The version is still in the 0.x range and moves quickly, so settings and UI may change
- The free plan’s ceiling (4 projects, 20 processes) is easy to hit if you juggle many repositories in parallel
Comparison with Similar Services
| Criteria | Solo | Warp | Conductor | tmux + iTerm2 |
|---|---|---|---|---|
| Positioning | Dashboard for processes and agents | AI-integrated development environment (ADE) | Isolated workspace per agent | General-purpose terminal multiplexing |
| Built-in AI | No (uses external CLI agents) | Yes (its own agent features) | Assumes Claude Code / Codex | No |
| Strength | Monitoring long-running processes, auto-restart, MCP integration | Unified terminal and agent experience | Parallel work and diff review via git worktrees | Highly flexible layouts |
| Platforms | macOS / Windows (Linux coming soon) | macOS / Windows / Linux | macOS | macOS / Linux |
| Pricing | Free tier; Pro at $99/year | Free tier / paid plans | Free tier / paid plans | Free (open source) |
Warp is moving toward rebuilding the terminal itself around AI, while Conductor isolates each agent in its own git worktree. Solo overlaps with neither: its focus is squarely on how you keep watch over the processes you are already running.
Who Is It For
- Developers who use Claude Code or Codex daily and always have five or more terminal tabs open
- Anyone running several long-lived processes — dev servers, workers, build watchers — where a silent crash is costly
- People running multiple agents in parallel who want to know which is idle and which has finished
- Those with a CLI-centric workflow who would rather not commit to a single AI vendor
- Teams that want a consistent environment startup procedure without adopting a heavy shared IDE
Summary
Solo consolidates everything that now runs at once in the age of AI agents into a single dashboard. Instead of bundling its own AI, it pairs with whichever CLI agent you already use and, as an MCP server, feeds process state back to that agent — which is where its distinctiveness lies. Since the free plan carries no feature restrictions, try it within the four-project limit to see whether it fits your workflow, and consider Pro once the number of repositories you handle grows.