Running Your Project

See an agent’s changes actually running — start a dev server, a build, or a script right inside its workspace, without leaving Fletch. That’s the Run panel: enable it under Settings › side panels › Run, then open the Run tab in an agent’s right rail.

Running your project

• Press Start to launch the detected command; it’s shown as $ <command>.
• Output streams into the panel as a live log.
• Press Stop to end the process.

When the process serves on a port, Fletch shows a :​<port> link that opens http://localhost:<port> in your browser.

Auto-detection

Fletch detects a sensible run command per ecosystem — Node, Python, Ruby, Rust, and Go — so most projects work with no setup. What it detects is grouped into Environment, Scripts, and Server settings.

Note

Each agent’s Run panel operates on that agent’s own checkout, so you can run several agents’ versions of the project side by side (mind the ports).

Your project’s .env isn’t visible inside the sandbox, so a run that needs secrets gets none by default. Share variables per project — mirrored live from .env, or overridden per agent — in project settings; see Run Configuration.

Related docs

• Run Configuration
• Terminal