🍴 Run your own radar

This project is a template for building an automated, AI-curated radar of new GitHub repositories in any niche you care about — a Rust radar, an AI-agents radar, a security radar. Setup takes about 10 minutes and runs itself weekly after that.

1. Fork the repository

Fork LokeshNanda/oss-radar-ai on GitHub.

2. Add your LLM API key

In your fork: Settings → Secrets and variables → Actions → New repository secret

• Name: OPENAI_API_KEY
• Value: your API key

Cost: roughly $0.05/week with the default gpt-4o-mini. You can also use any OpenAI-compatible provider — OpenRouter, Groq, or a self-hosted Ollama — by setting OPENAI_BASE_URL and OPENAI_MODEL (see step 4).

3. Enable GitHub Pages

Settings → Pages → Build and deployment → Source: GitHub Actions

4. Scope your radar (optional)

Add repository variables (Settings → Secrets and variables → Actions → Variables tab) to customize your radar:

Variable	Example	Effect
RADAR_SEARCH_QUERY	topic:rust language:rust	Only track new Rust repos
RADAR_SITE_NAME	Rust Radar	Site title
RADAR_SITE_DESCRIPTION	New Rust repos, analyzed weekly	Site description / SEO
RADAR_SITE_URL	https://you.github.io/oss-radar-ai/	Used in feeds and links

More query ideas: topic:ai language:python, topic:security, topic:kubernetes, stars:>100 topic:llm. Any GitHub search qualifier works.

Then wire them into the workflow's env block in .github/workflows/trending.yml:

      - name: Run radar pipeline
        env:
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          RADAR_LOG_LEVEL: INFO
          RADAR_SEARCH_QUERY: ${{ vars.RADAR_SEARCH_QUERY }}
          RADAR_SITE_NAME: ${{ vars.RADAR_SITE_NAME }}
          RADAR_SITE_DESCRIPTION: ${{ vars.RADAR_SITE_DESCRIPTION }}
          RADAR_SITE_URL: ${{ vars.RADAR_SITE_URL }}
        run: |
          radar-run

      - name: Build MkDocs site
        env:
          RADAR_SITE_NAME: ${{ vars.RADAR_SITE_NAME }}
          RADAR_SITE_DESCRIPTION: ${{ vars.RADAR_SITE_DESCRIPTION }}
        run: |
          mkdocs build --strict

Also reset the state and reports from the original radar so yours starts fresh:

rm -f .radar_state/*.json docs/repos/*.md docs/reports/*.md docs/categories/*.md

Using a different LLM provider

Set these as repository secrets/variables and pass them into the pipeline env:

# OpenRouter
OPENAI_BASE_URL=https://openrouter.ai/api/v1
OPENAI_MODEL=openai/gpt-4o-mini

# Groq
OPENAI_BASE_URL=https://api.groq.com/openai/v1
OPENAI_MODEL=llama-3.3-70b-versatile

5. Run it

Actions → Open Source Radar → Run workflow. Your radar site publishes to https://<you>.github.io/oss-radar-ai/ and re-runs automatically every Monday.

---

Questions or improvements? Open an issue or see CONTRIBUTING.