otter
First run

Self-host

Run Otter on your machine in a few minutes. No need to clone the repo unless you want to contribute — pull the public image or install the CLI from npm.

Install

docker pull manasdutta04/otter

Option A — Docker (full UI)

  1. Pull: docker pull manasdutta04/otter
  2. Start Compose (creates local Postgres + Redis):

Canonical compose URL: docker compose -f https://otter.manasdutta.com/docker-compose.yml up -d

Fallback (same host as this docs build): docker compose -f https://otter-2ookq06sf-manas-projects-577f14e5.vercel.app/docker-compose.yml up -d

Open http://127.0.0.1:3000/app. Image: manasdutta04/otter:latest.

Requirements

  • Docker Engine + Compose v2
  • Ollama on the host (recommended): ollama pull qwen2.5-coder:7b

Windows note

If Compose fails resolving .env from a remote URL, download docker-compose.yml first, then run docker compose -f docker-compose.yml up -d.

Option B — CLI only

npm i -g @otter-engg/cli
otter

No Docker. Storage under ~/.otter/. See CLI docs for slash commands. Same package for pnpm / yarn / bun.

Connect GitHub

In the Docker UI click Connect GitHub, or in the CLI run otter login / /login. You install the Otter GitHub App — you do not paste Client secrets. Details: GitHub docs.

Open the product UI (Docker)

  • Workspace: http://127.0.0.1:3000/app
  • Models: http://127.0.0.1:3000/app/models
  • API: http://127.0.0.1:8000

First-run checklist

Prefer the step-by-step guide: First 10 minutes.

  1. Install via Docker pull or npm CLI (@otter-engg/cli).
  2. Connect GitHub (Otter app).
  3. Point models at Local Ollama (or OpenAI-compatible).
  4. Import a repository.
  5. Ask → plan → approve before any write.

Want to hack on Otter?

Clone the public repo and see Contribute.