
1Code is an open source Claude Code client that provides a calm and visual interface for parallel agent work. It is built for developers who rely on AI coding agents like Claude Code and Codex, offering a unified dashboard to manage them. The core value lies in its ability to let users run multiple coding agents concurrently, each in its own session with full context, significantly accelerating feature development and codebase exploration. By combining the power of state-of-the-art AI models with a thoughtfully designed user interface, 1Code transforms how developers interact with coding agents, moving from terminal-driven workflows to a visual, keyboard-first experience. It is backed by Y Combinator and is available both as a self-hosted open source version and a paid Pro tier with additional capabilities.
The primary pain point 1Code addresses is the fragmentation and inefficiency inherent in using multiple coding agents from the command line. Developers previously had to switch between separate terminal windows, manage different API keys, and struggle with context loss when moving between agents. This fragmented workflow made it nearly impossible to work on different features simultaneously with agents, leading to bottlenecks and wasted time. 1Code eliminates these issues by providing a single application that houses both Claude Code and Codex, allowing instant switching and parallel execution. This matters because it directly impacts developer productivity, enabling teams to ship features faster by leveraging the unique strengths of each agent without the overhead of managing multiple environments.
The first major feature group is Parallel Agents, which allows users to work on multiple features at once. Each agent runs in its own independent session, maintaining its own context and chat history. This means a developer can assign one agent to refactor a function while another investigates a bug in a completely different part of the codebase, all within the same app. The benefit is a dramatic reduction in context switching and waiting time. Instead of queuing tasks for a single agent or managing separate terminals, users can orchestrate multiple agents simultaneously. This feature is especially valuable during large refactors or when tackling interdependent features, as it provides a clear overview of all active agent activities from a single pane.
The second major feature group is Git Integration, which is built end-to-end into 1Code. It includes Visual Staging & Diffs, allowing developers to see exactly what changed, stage individual files, review diffs with a graphical diff view, and commit changes with confidence. PR Creation enables opening pull requests directly from the UI, complete with descriptions, reviewers, and labels. Most importantly, Worktree Isolation ensures each agent session runs in its own Git worktree, preventing file conflicts and eliminating the need for stashing. This makes it safe to run multiple agents making changes to the same repository simultaneously. The integration streamlines the entire Git workflow from code generation to merge, keeping everything within the calm interface.
admin
The third major feature group encompasses Background Agents and MCP Integration. Background Agents run in isolated cloud sandboxes that keep working even when the laptop is closed, providing live browser previews of development branches. Users can toggle between desktop and mobile viewports to see live changes. MCP Integration allows connecting any MCP-compatible server, such as Notion, Linear, GitHub, Slack, Sentry, or PostgreSQL, enabling agents to read errors, manage tasks, update documentation, and more. Additionally, @1code triggers allow tagging agents directly from GitHub, Linear, or Slack to start a task on demand. Automations can auto-review PRs, fix CI failures, and complete Linear tasks, all configured within the app. These features extend 1Code from a simple client into a powerful automation hub.
Overall, 1Code follows a three-step workflow to ship code. First, users connect an agent by signing in with Anthropic, OpenAI, or by bringing their own API key. Second, they can run agents anywhere—from the desktop app, web, or mobile PWA (on the Pro plan). Third, they ship directly from the UI by opening PRs, reviewing diffs, and merging—all without leaving the application. The design philosophy is keyboard-first, with hotkeys for every action, allowing navigation, creation, and management without touching a mouse. The open source nature invites community contributions and self-hosting, while the Pro tier adds background execution, live browser previews, and the ability to start agents from a phone or web browser.
Concrete use cases include a developer working on two features concurrently: one agent refactors the authentication module while another updates the UI for a dashboard, both operating on separate worktrees without conflict. Another scenario is a team using background agents to run test suites and fix bugs overnight—the agents keep working in cloud sandboxes, and in the morning the developer reviews PRs automatically opened from the fixes. A third scenario involves a product manager tagging @1code in a Linear task description to spawn an agent that implements the described feature, then watches a live browser preview of the branch. These use cases demonstrate how 1Code enables continuous coding and review cycles that extend beyond the developer's active work hours.
1Code is designed for individual developers and engineering teams who already use AI coding agents like Claude Code or Codex and want a more efficient, visual, and parallel workflow. The open source plan supports macOS, Windows, and Linux with community support and full source code access. The Pro plan at $20 per month adds desktop, web, and mobile access, background agents, live browser previews, and the ability to start agents from any device. The platform leverages the MCP protocol for extensibility and is backed by Y Combinator. In summary, 1Code elevates the coding agent experience by providing a calm, visual client that maximizes parallel execution, minimizes context switching, and integrates deeply with development workflows, making it an essential tool for any developer serious about AI-assisted coding.
Software developers and engineering teams who use AI coding agents like Claude Code and Codex and need a streamlined, visual interface for parallel execution. Suitable for solo developers working on multiple features simultaneously, small teams looking to reduce context switching, and larger organizations that want to integrate agent-based workflows with existing tools via MCP. The open source option appeals to self-hosting enthusiasts, while the Pro plan targets power users who require background agents, live browser previews, and mobile access. Ideal for those who value a keyboard-first, calm UI and automated Git workflows.