Emdash: Open-Source Agentic Development Environment for Parallel Coding Agents with Multi-Provider Support and Remote SSH Workflows
Emdash is an open-source agentic development environment designed for modern software development. Built for the YC W26 cohort, it enables users to run multiple coding agents in parallel, each operating in its own isolated Git worktree to maintain clean and organized changes. This setup allows developers to work on multiple features simultaneously, with each agent handling a specific task independently. Emdash is provider-agnostic, supporting over 15 CLI-based AI coding agents such as Claude Code, Qwen Code, Amp, Codex, GitHub Copilot, Gemini, Mistral Vibe, and more. Users can easily install and configure these agents via standard package managers like npm, curl, or homebrew. The environment also supports seamless handoff of tickets from project management tools like Linear, GitHub, and Jira directly to agents, with side-by-side diff reviews for easy code validation. For teams working with remote codebases, Emdash offers full SSH and SFTP integration. Developers can connect to remote servers securely using SSH agent or private key authentication, with credentials stored safely in the operating system’s keychain. Once connected, users can run agents on remote projects using the same parallel workflow as local development, making it ideal for distributed teams and cloud-based workflows. The application is designed with privacy in mind. All app data is stored locally in a SQLite database, and no code, file paths, prompts, or personally identifiable information is sent to external servers by Emdash itself. However, when using third-party agents, data is transmitted to the respective provider’s cloud API for processing—each provider has its own data handling and retention policies. Telemetry is optional and minimal. Emdash collects anonymous, allow-listed events such as app start and close times, feature usage, and version information, which are sent to PostHog. Users can disable telemetry at any time via configuration. Installation is straightforward across platforms: macOS, Windows, and Linux. Users on macOS can install via Homebrew, while others can use package managers or direct installation scripts. For remote development, detailed setup guides are available in the documentation. Contributions are welcome. The project encourages developers to add new providers, improve documentation, or report issues. A Discord community is available for discussion and support. In summary, Emdash empowers developers to scale their coding workflows with AI agents, maintain clean Git histories, collaborate across remote environments, and retain full control over their data—all while working with a growing ecosystem of AI coding tools.
