Grep Badger logoGrep Badger
← Back to docs

Getting Started

Grep Badger is built around hosts. In this app, a host is any machine or server you connect to and read logs from — your local PC via localhost, or a remote system over SSH.

1. Add a host

Grep Badger starts with a local host already available.

To add a remote SSH host:

  1. Click the + icon to the right of Files in the left sidebar. This opens the file explorer.
  2. Click the + icon to the right of Hosts in the left sidebar of the file explorer.
  3. Fill out the necessary details, then test and save.

Windows and Linux are currently supported. On Linux, passwords may not be remembered if a credential store is unavailable. If the app cannot save your password, it will prompt you when you connect. It will hold your password in memory for your session only and prompt you again next session.

2. Open a log file

  1. Click the + icon to the right of Files in the left sidebar.
  2. In the file explorer, select your host.
  3. Navigate to the log file you want to open.

Once you select a log file, Grep Badger opens it in the main view and begins tailing the file with a clear history.

Remote log files are streamed directly from the host over SSH/SFTP. They are not copied locally.

For large files, this may take a moment depending on file size and network speed.

3. Tail the log

After the log is open, Grep Badger will stream new lines as they arrive. You can pause and start the tail at any time by clicking the Pause/Tail button in the CONTROLS section in the left sidebar.

Use this when you want to watch live output from:

  • application logs
  • system logs
  • container logs
  • any other streamed text source

4. Loading chunks

Grep Badger streams files instead of loading them directly into memory. This allows Grep Badger to minimize its memory footprint. To maintain quick performance, navigation is done in chunks.

Chunk navigation is done in the CONTROLS menu in the left sidebar. You can Load Latest, Previous, Next or Go to a line.

Due to streaming, line numbers are estimated. When navigating to a line, treat it as a ballpark rather than a definitive position.

When you do any chunk navigation, tailing will be paused. When you start tailing again, the latest chunk will always be loaded first.

The default chunk size is 1000. This is configurable in settings.

5. Save open log to Quick Access

If you plan to open the same log again later, save it to Quick Access.

Quick Access is for log files you want to reopen quickly without navigating the file explorer again.

Tabs do not persist after closing the app.

6. Using AI

The right panel contains an AI interface. By default, Ollama is created as a provider. You can select context to send with your AI request by clicking lines within the log viewer.

For privacy, we recommend using a local AI model when possible.

The app assumes Ollama is installed and a model is pulled. If this is not the case, Ollama will not be usable. Please install Ollama if you wish to use local AI.

There are many models capable of running on your CPU. While a GPU would help, you do not need a large GPU to run local models.

Important notes

  • Remote log files are streamed through SSH/SFTP. Very large files may affect performance.
  • External AI providers will see any data you send to them. Many providers use your chats for training and may hold onto your history. Never send secrets to an external AI provider. For this reason we highly recommend using a local AI provider such as Ollama.