vibn-agent-runner/vibn-dev/README.md

vibn-dev

Per-project AI development container. One of these runs in Coolify per Vibn project; the AI agent drives it via shell.exec and fs.* MCP tools.

See /AI_PATH_B_EXECUTION_PLAN.md for the architecture.

Build & publish

docker buildx build --platform linux/amd64 -t ghcr.io/vibnai/vibn-dev:latest --push .

The image is pre-pulled on every Coolify host on deploy so first-use spin-up stays under 5 seconds.

Smoke test locally (requires Docker with QEMU for cross-platform)

docker buildx build --platform linux/amd64 -t vibn-dev --load .
docker run --rm -it -v "$PWD/scratch:/workspace" vibn-dev bash
# inside: node --version    # Node LTS ships in the image
# inside: npm --version
# inside: python3 --version
# inside: go version
# inside: rg --version      # ripgrep ships in the image
# inside: git --version

What's in the image

• Node.js LTS (via NodeSource)
• Python 3.12 + pip + venv
• Go 1.23
• ripgrep, git, build-essential, jq, curl, nano, vim
• supervisord (reserved for future dev_server use)
• tini (PID 1 reaper)

What's NOT in the image (by design)

• Coolify control-plane creds — never. The container has no route to internal Vibn services (Docker network policy enforced at host level)
• SSH server — exec happens via docker exec from the Coolify host