Your personal AI Software Engineer
587
Self-hosted engineering intelligence: understand repositories, chat with citations, plan changes, and ship approval-gated patches — with your own local model (Ollama) or any OpenAI-compatible endpoint.
Image: manasdutta04/otter
Tags: latest (main), 0.1.0 (release), sha-<commit> (traceability)
# clone the repo (compose + env example)
git clone https://github.com/manasdutta04/otter.git
cd otter
cp .env.example .env
# set GITHUB_CLIENT_ID / GITHUB_CLIENT_SECRET
# OAuth callback: http://127.0.0.1:8000/auth/github/callback
docker compose -f docker/compose.platform.yml pull
docker compose -f docker/compose.platform.yml up -d
Open:
One platform container:
:3000):8000)Compose also starts official Postgres 16 and Redis 7. Ollama stays on the host.
docker pull manasdutta04/otter:latest
# or pin a release
docker pull manasdutta04/otter:0.1.0
ollama pull qwen2.5-coder:7b
From Docker, Otter reaches the host via http://host.docker.internal:11434/v1 (PLATFORM_LLM_BASE_URL).
/docs on the marketing site)Content type
Image
Digest
sha256:851927fb1…
Size
162.7 MB
Last updated
about 2 months ago
docker pull manasdutta04/otter