EZ PM2 GUI: a web dashboard for managing PM2 and investigating logs
I build and maintain EZ PM2 GUI, an open-source web interface for the PM2 process manager. It brings process controls, metrics and logs together in a browser, including when you’re working with remote servers. The current release, v1.13.0, expands the log investigation workflow. Here’s a look at what the project does and how to get started. From a process list to an investigation The dashboard shows PM2 processes alongside CPU and memory information. You can start, stop and restart processes, inspect logs, view live and historical metrics, and connect to remote servers over SSH. There are also interfaces for deployment, cluster management, cron jobs and PM2 modules. The aim is to make common operational tasks accessible from one place while keeping PM2 underneath as the process manager. AI-assisted log analysis The log viewer can send a bounded snapshot of local or remote logs for analysis using an installed Claude Code, Codex, Cursor Agent or GitHub Copilot C
I build and maintain EZ PM2 GUI, an open-source web interface for the PM2 process manager. It brings process controls, metrics and logs together in a browser, including when you’re working with remote servers.
The current release, v1.13.0, expands the log investigation workflow. Here’s a look at what the project does and how to get started.
From a process list to an investigation
The dashboard shows PM2 processes alongside CPU and memory information. You can start, stop and restart processes, inspect logs, view live and historical metrics, and connect to remote servers over SSH. There are also interfaces for deployment, cluster management, cron jobs and PM2 modules.
The aim is to make common operational tasks accessible from one place while keeping PM2 underneath as the process manager.
AI-assisted log analysis
The log viewer can send a bounded snapshot of local or remote logs for analysis using an installed Claude Code, Codex, Cursor Agent or GitHub Copilot CLI, or an Ollama model. The CLI tools use their existing sign-in; EZ PM2 GUI does not bundle the agents, models or credentials.
Before analysis, you can review the log snapshot. Common credentials are redacted, and cloud analysis requires explicit consent. Redaction is not a guarantee that every sensitive value is removed, so review the snapshot before sending it. The report suggests diagnostic checks; it does not make process changes for you.
In v1.13.0, investigations have a persistent sidebar with separate conversations, follow-up questions and SQLite-backed history. Analysis can continue while the sidebar is hidden or the page is reloaded. You can also choose local or cloud Ollama models and optional model overrides for supported installed CLI providers.
The earlier AI log analysis demo shows the preview and consent flow using sample logs and a simulated report. There’s also a product walkthrough.
Try it
You’ll need Node.js and PM2 installed. Install the GUI globally and start it:
npm install -g pm2
npm install -g [email protected]
ezpm2gui
Then open http://localhost:3101. If your processes use a custom PM2 home, set PM2_HOME to the same directory before starting the GUI. The README covers configuration, remote connections and the provider setup in more detail.
The project is open source under AGPL-3.0-or-later. If you use PM2, I’d love to hear which parts of your monitoring or debugging workflow could be easier.
