Quickstart

1. Install

With Homebrew:

brew install openotters/tap/otters
brew services start otters

Or with Go:

go install github.com/openotters/openotters/cmd/otters@latest
go install github.com/openotters/openotters/cmd/ottersd@latest
ottersd serve &

2. Configure a provider

otters provider add

This walks you through wiring an LLM provider — Anthropic, OpenAI, Ollama, and others. The provider config is stored locally; the daemon never round-trips your keys to a SaaS.

3. Run a demo agent

otters run ghcr.io/openotters/agents/pinger:latest --name pinger
otters chat pinger

That pulls the public image, starts the agent in your daemon, and drops you into a chat session.

4. One-shot prompts

$ otters prompt pinger "google.com"
google.com: reachable

Same agent, different surface — useful in scripts, CI, and pipelines.

5. Open the dashboard

The daemon ships a browser UI on the same port. otters info prints the URL (default http://127.0.0.1:5500). The dashboard lets you:

• Browse the image catalog and launch an agent with MODEL + every ENV overridable in one dialog — no Agentfile edit required.
• Watch async jobs stream live stdout / stderr on /jobs/<id>.
• Chat with any agent (sessions are shared with the CLI — start in the TUI, finish in the GUI, or vice versa).

Agents that need configuration (auth tokens, base URLs) are easiest to spin up from the dashboard: pick the image card, hit Run, fill the ENV overrides in the dialog, and you're chatting.

What's next

• Build your own agent — write an Agentfile, push the image.
• CLI reference — every command in one place.