OwlSight

Appearance

Docker ​

OwlSight ships with a Dockerfile for containerized usage. The image includes the .NET runtime and Git.

Build ​

bash
cd OwlSight
docker build -t owlsight .

The multi-stage Dockerfile:

  1. Build stage — Uses dotnet/sdk:10.0-preview to restore and publish
  2. Runtime stage — Uses dotnet/runtime:10.0-preview with Git installed

Usage ​

Mount your repository at /repo and pass review options as arguments:

bash
docker run --rm \
  -v $(pwd):/repo \
  -e OWLSIGHT_API_KEY=$OPENAI_API_KEY \
  owlsight review --base main

With JSON Output ​

bash
docker run --rm \
  -v $(pwd):/repo \
  -e OWLSIGHT_API_KEY=$OPENAI_API_KEY \
  owlsight review --base main --output /repo/report.json

The report is written inside the mounted volume, so it persists after the container exits.

With a Local LLM ​

If your LLM runs on the host machine, use host.docker.internal (macOS/Windows) or --network host (Linux):

bash
# macOS / Windows
docker run --rm \
  -v $(pwd):/repo \
  owlsight review --base main \
  --base-url http://host.docker.internal:11434/v1 \
  --api-key ollama \
  --model llama3

# Linux
docker run --rm --network host \
  -v $(pwd):/repo \
  owlsight review --base main \
  --base-url http://localhost:11434/v1 \
  --api-key ollama \
  --model llama3

Initialize Config ​

bash
docker run --rm \
  -v $(pwd):/repo \
  owlsight init

CI/CD with Docker ​

See the CI/CD Integration guide for pipeline-specific examples using Docker.

yaml
# GitHub Actions example
- name: Run Review
  run: |
    docker run --rm \
      -v ${{ github.workspace }}:/repo \
      -e OWLSIGHT_API_KEY=${{ secrets.OPENAI_API_KEY }} \
      owlsight review --base origin/main --output /repo/report.json

Environment Variables ​

Pass environment variables with -e:

bash
docker run --rm \
  -v $(pwd):/repo \
  -e OWLSIGHT_API_KEY=sk-... \
  -e OWLSIGHT_MODEL=gpt-4o-mini \
  -e OWLSIGHT_BASE_URL=https://api.openai.com/v1 \
  owlsight review --base main

See Configuration for all supported variables.