Getting Started¶
In this tutorial, you'll install VlinderCLI, verify the setup, and run the bundled support agent.
Time: ~10 minutes
Prerequisites: macOS or Linux, and a terminal.
Step 1: Install VlinderCLI¶
The script downloads the vlinder binary, creates ~/.vlinder/ with a default config, and checks for prerequisites.
If NATS or Podman aren't installed, the script stops and prints the install commands you need. Install them, then re-run the script.
See Installation for full details, prerequisites, and platform-specific instructions.
Step 2: Verify¶
After a successful install, NATS and the vlinder daemon are running as system services. Verify:
This connects to the daemon and shows deployed agents. You should see the support fleet's agents listed.
Step 3: Run the Support Agent¶
VlinderCLI ships with a bundled support fleet. Try it:
You'll see an interactive REPL. Ask it a question about VlinderCLI and press Enter.
Tip
Type exit or press Ctrl+C to end the session.
Step 4: View the Timeline¶
After your conversation, inspect what happened:
Each entry shows a commit SHA, timestamp, and agent name — every interaction is recorded in a git-backed timeline.
What Just Happened?¶
- The vlinder daemon (already running as a service) received your request
- The support fleet's entry agent loaded and started in a container
- Your inputs were routed through NATS to the agent runtime
- Each interaction was committed to the timeline at
~/.vlinder/conversations/
Next Steps¶
- Your First Agent — scaffold and run your own agent
- Configuration — customize logging and providers
- Architecture — understand the component model