Your AI coding assistants
You use AI coding assistants — probably several. 1DevTool is the desktop workspace that organizes them all: Claude Code, OpenAI Codex CLI, Gemini CLI, Amp, Aider, and any CLI-based assistant. Named persistent terminals, 2×2 grid layouts, and built-in HTTP + database tools so you can test what your assistants write without leaving the window.
$29 one-time after beta. macOS, Windows, Linux.
Sound familiar?
AI assistants are scattered across terminals
Claude Code in one iTerm tab, Codex CLI in another, Gemini CLI somewhere else. No unified view, no shared context, no way to run them in parallel.
Switching costs are real
Every time you switch between AI assistants or between projects, you lose context, lose the conversation, and restart from zero.
No project-to-assistant memory
Which AI assistant did you use for which project last week? No tab, no history, no session restore — just lost work.
Can't test what your assistant writes
Your assistant generates an API route. Now you pivot to Postman, then a DB client, then Chrome — and your assistant sees none of the results.
Crash = all sessions gone
A reboot or terminal crash kills every running assistant session. Re-auth, re-open the repo, re-prompt.
One workspace for all your AI assistants
- All assistants in named, color-coded terminals — 'Claude Frontend', 'Codex API', 'Gemini Docs'
- Per-project workspaces keep assistant sessions scoped and organized
- 2×2 grid layouts — watch Claude, Codex, and Gemini work in parallel
- Tmux-backed persistent sessions on macOS/Linux survive restarts
- Built-in HTTP client, 13-engine database client, embedded Chromium browser — test what assistants write
- Send-to-AI sends screenshots, HTTP responses, and DB rows straight into your assistant's input
Everything you need for ai coding assistant
Multi-Assistant Dashboard
Claude Code, Codex CLI, Gemini CLI — all visible at once in a 2×2 grid. Agent state (running, waiting, completed) shown in the Terminal Dashboard.
Named Sessions
Name your AI assistants. 'Claude — Frontend', 'Codex — API', 'Gemini — Docs'. Color-coded so you always know which agent is which.
Project Workspaces
Each project remembers which AI assistants you were using, which terminals you had open, and which files you were editing.
Instant Switching
Switch projects in one click. All your AI assistants are exactly where you left them — no re-prompting, no re-auth.
Integrated HTTP + DB + Browser
Test the APIs, data, and UI your AI assistants generate. Send results back into the assistant via Send-to-AI.
Works With Any Assistant
Not locked to one provider. Run Anthropic Claude Code, OpenAI Codex CLI, Google Gemini CLI, Sourcegraph Amp, Aider, or your own tool.
Frequently asked questions
What's the best workspace for AI coding assistants?
For running multiple CLI-based AI coding assistants (Claude Code, Codex CLI, Gemini CLI) in parallel, 1DevTool is purpose-built — named persistent terminals, 2×2 grids, Send-to-AI, and an integrated HTTP + DB + browser toolchain. For single-model inline completion inside VS Code, Copilot or Cursor are closer fits.
Which AI coding assistants does 1DevTool support?
Any CLI-based assistant: Anthropic Claude Code, OpenAI Codex CLI, Google Gemini CLI, Sourcegraph Amp, OpenCode, Cline, Qoder, Aider, Continue, and any custom agent. Local models via Ollama and LM Studio work as custom agents.
Is this like GitHub Copilot?
Different category. Copilot does inline completion inside your editor. 1DevTool is a workspace for running full AI agents — Claude Code, Codex, Gemini — that plan, edit, and test autonomously. You can pair Copilot inside Monaco with 1DevTool's agents if you want both.
Do I need to install the AI assistants separately?
Yes. Install Claude Code, Codex CLI, Gemini CLI, etc. separately via their official instructions. 1DevTool is the workspace that runs and organizes them — it does not ship the agents themselves.
Can I run multiple assistants in parallel?
Yes. Put Claude Code, Codex CLI, and Gemini CLI in a 2×2 grid and watch them work simultaneously on different parts of the same repo. Agent Pipelines let you chain them automatically (planner → implementer → reviewer).
Do my sessions survive restarts?
Yes, on macOS and Linux — terminal sessions are tmux-backed so they survive restarts, crashes, and closed laptops. On Windows, sessions restore on app relaunch.
What platforms does it run on?
Native builds for macOS (Apple Silicon + Intel), Windows 10/11, and Linux (.deb + AppImage).
How much does it cost?
Free tier with 2 projects and 500 lifetime AI messages. $29 one-time purchase for unlimited use. No subscription. AI API usage billed separately by each provider (Anthropic, OpenAI, Google).
Organize your AI coding assistants
Every AI. Every project. One workspace.
Download 1DevTool — FreemacOS available now. Windows & Linux coming soon.