Runlist

Every local app, across every repository, in one focused VS Code control panel.

Save each project's folder and commands once, then start, stop, inspect, and open it from the sidebar.

Set projects up yourself—or optionally let a supported coding agent propose the setup for your approval. Runlist checks configured ports before start and never stops an unknown process to free one.

A control panel for local projects

Runlist keeps a reusable list of local projects from different repositories inside VS Code. It combines saved commands with project status, cautious port checks, quick links, and recent output, so it is more than a list of tasks to run.

• Start projects with their saved commands and safely stop the process trees Runlist launched.
• Search saved projects by name or folder.
• Pin important projects so they stay at the top of the list.
• Give projects a friendly name without renaming their folders.
• Keep configured service names and local addresses visible at a glance.
• When another Runlist-owned project needs the same port, switch safely by stopping it before starting the selected project.
• When service ports are configured, see project status, copy a responding web service's URL, and open the first service at its localhost address or optional HTTP/HTTPS URL. Services with an Open URL are also checked for an HTTP response.
• Expand a running web app into stable Overview, Output, Preview, and History tabs, with refresh, copy, and browser actions.
• Open an eligible responding local web app on a phone by scanning a QR code generated entirely on your computer.
• See current CPU and memory use for the process tree Runlist started while its project details are expanded.
• Open any saved project folder in a new VS Code window.
• Expand a starting or running project to follow its startup timeline and the latest three useful output lines, then open the complete Recent Output screen when you need more detail.
• View readable recent output from the latest run, including a concise failure summary when startup exits, follow new lines, and open web links without leaving the sidebar.
• After a failed start, prepare a bounded, sanitized diagnosis request for a connected coding agent without sending anything automatically.
• Edit or remove a saved project without touching its files.

Runlist remembers the setup. You decide what runs.

Install

Runlist is tested on Windows, macOS, and Linux.

Install from the VS Code Marketplace

Install Runlist from the VS Code Marketplace, or install it from VS Code:

1. Open the Extensions view in VS Code.
2. Search for Runlist.
3. Confirm that the publisher is Hanko Swart and the extension ID is hankoswart.runlist.
4. Select Install.

After installation, select the Runlist icon in the VS Code Activity Bar.

Add your first project

1. Select the + button in the Runlist sidebar.
2. Optionally enter a friendly project name. If you leave it blank, Runlist uses the folder name.
3. If a local folder is already open in VS Code, choose Use current workspace, or select Browse. In a multi-root workspace, Runlist asks which local workspace folder to use.
4. Enter the command that starts the project.
5. If the project needs a special shutdown workflow, optionally enter a custom stop command. Most projects should leave this blank.
6. If you know them, add each service name and port so Runlist can verify its status. You can optionally add an HTTP or HTTPS URL to open instead of the service's localhost address. Adding this Open URL also lets Runlist check whether the web service is responding.
7. Save it.

Runlist points out missing or invalid details beside the field that needs attention. If you close the screen after making changes, it asks before discarding them.

Your project is now ready whenever you need it. Select the Start icon to run its saved start command. Projects with configured services stay Starting… until every saved TCP port is accepting connections. A service with an Open URL also waits for its web page to respond; redirects, sign-in responses, and error pages still count as a response. After 30 seconds, Runlist shows Taking longer… and names the services it is still checking, without stopping the project or giving up. If those services become ready later, the status changes to Running. A service whose port is open but whose page does not respond is shown separately as Web service not responding. Services without an Open URL keep using port-based checks. Projects without configured services use the launched process state. Select Stop to stop only the process tree Runlist launched. When an explicit custom stop command is configured, Runlist runs it instead.

If Runlist is open in more than one VS Code window, starting or stopping a project in one window updates its status in the others automatically.

Configured service ports are lightweight details, not a port-management system. Projects may save the same service ports because they can still run at different times. Runlist points this out while you add or edit a project, checks every configured port again before starting, and blocks the start if another app is already using one.

If another Runlist project owns a port, Runlist names it. If a shared port is occupied but its owner cannot be identified safely, both Start and Stop remain unavailable until that port is free. Runlist never changes ports or stops an unknown process automatically.

You can set up every project yourself. If you prefer, the optional coding-agent setup below can inspect a project and propose its commands, service ports, and explicitly documented browser URLs for your approval.

Optional: set up a project with your coding agent

Runlist can also give a supported coding agent a guided setup skill. The agent inspects the project, finds its existing commands, service ports, and explicit browser URLs, and saves the result through Runlist.

1. Select the plug button beside the + button to open Agent connections.
2. Select Set up beside the agent you use.
3. Restart Codex after setup. If an already-open Claude Code session does not find the new skill, restart it too.

Then open the project with your agent and use its Runlist skill:

You can also describe what you want naturally. For example:

Inspect this project and add it to Runlist with the name My App. Identify the exact start command and the port for every service it runs. Preserve any explicit HTTP or HTTPS browser URLs, and include a custom stop command only if the project daemonizes or manages external services such as Docker or databases.

The agent can propose a new project or an update to one already in Runlist. The project then shows Review setup in the sidebar. Check its folder and exact commands, then select Approve setup before Start or Stop becomes available. Agent-proposed commands never run without this approval.

When a start fails, open View output and select Ask your agent. Runlist first shows exactly which saved details and retained output the agent can retrieve. Select Copy diagnosis request, then paste it into your agent chat. Nothing is sent automatically, and any setup change the agent proposes still returns to Runlist for your review and approval.

After updating Runlist, return to Agent connections and select Refresh setup so the agent uses the current connection and skill. Runlist will not replace a different skill you created with the same name.

Day-to-day use

Runlist keeps the everyday controls simple:

Removing a project from Runlist does not delete the project or any of its files.

If Runlist finds a configured service already running but did not start it itself, the project is labelled Detected running so its state is clear.

Your projects stay local

Runlist stores its project list in your local VS Code data. Runlist itself does not upload project folders, commands, service ports, or URL overrides. Any coding agent you connect has its own data and privacy settings.

Security

Found a possible security issue? Please read the security policy and report it privately. Do not share sensitive details in a public issue.

Runlist is available under the MIT License.