The questions people actually ask, answered plainly. If yours isn’t here, the troubleshooting page covers things that go wrong, and the in-app email button reaches a real human.
Does Termic work with my Claude Pro / Max subscription?
Yes. Termic spawns the interactive claude CLI, the same binary you’d run in iTerm, with the same auth and the same quota. Anthropic’s June 15, 2026 Agent SDK credit change is scheduled to apply to the SDK path and claude -p; interactive Claude Code stays on the regular subscription usage limits. Termic rides that. See Introduction.
Will I burn through my plan faster with Termic?
You’ll spend the same quota you’d spend running claude in a terminal, because that’s literally what Termic runs. There’s no SDK wrapper and no per-token markup from Termic. Running several agents in parallel does spend more (each is a real session), but each one draws from your normal subscription, not a separate metered pool.
How is this different from Conductor?
The honest differences are source, platform support, agent breadth and process boundaries. Termic is AGPL-3.0, ships macOS and Linux builds, supports Grok, Antigravity, Copilot, opencode and custom PTY agents, has main checkout and multi-repo task modes, and can put each task behind a macOS Seatbelt sandbox plus network allowlist. Full breakdown: Migrating from Conductor and the comparison page.
Does Termic see my prompts or my code?
No. Termic is a local app with no backend. Prompts go from the terminal to the CLI to that CLI’s vendor, exactly as they would without Termic; Termic doesn’t read or forward them. See Privacy.
Can an agent touch my secrets or hit random servers?
Not if you enable the sandbox on the task. macOS Seatbelt blocks reads and writes to ~/.ssh, ~/.aws, Keychains, browser profiles and your personal folders, and network goes through an allowlist of hosts you approve. The sandbox is pinned at task creation, so an agent can’t disable the cage it was spawned in.
Which agents does it support?
claude, codex, agy (Antigravity), copilot, grok and opencode are built in and auto-detected. You can register any other terminal-based agent (aider, ollama, a shell script) in about thirty seconds. See Agents & the registry.
Is it really open source?
Yes, AGPL-3.0, on GitHub. Fork it, audit it, package it. The license means derivatives stay open too, which is what prevents the “open core quietly goes proprietary” pattern.
Does it run on Linux and Windows?
Linux ships as a signed AppImage with every release, with the same in-app updater as macOS. Windows builds from source today; prebuilt installers are on the roadmap. The sandbox is macOS-only for now on both. See Install.
Do I have to use worktrees for everything?
No, and that’s a deliberate difference from Conductor. You can attach an agent straight to your main checkout for quick changes that don’t deserve a branch, and only spin up a worktree when you actually want isolation or parallelism. See Tasks & worktrees.
How big is it and how fast does it launch?
Around 10 MB, and it launches instantly. It’s a Tauri app (Rust + the system WebView), not Electron. See Architecture & data.
Related
- Introduction: the subscription / SDK story in full.
- Troubleshooting: fixes for things that go wrong.
- vs Conductor: the detailed comparison.
- Best Conductor alternative: the broader alternatives guide.