Open Source Radar

Find an open source issue you can start today, and learn how to turn it into a merged pull request.

Open, unclaimed, newcomer-friendly issues from active open source projects, sorted by language and topic and refreshed every 12 hours, plus a complete, practical contribution guide.

Browse issues on the website  |  Read the guide  |  Issues by language  |  Issues by topic  |  Projects directory  |  Dataset on Hugging Face

Found it useful? A star helps other newcomers find the radar.

Right now on the radar

3,381 open issues · 1,287 labeled for beginners · 1,174 active projects · updated 2026-10-06 22:38 UTC

Languages: C++ (352) · Go (326) · C# (319) · Rust (299) · TypeScript (278) · Java (263) · Python (233) · Kotlin (158) · PHP (157) · C (155) · JavaScript (150) · Shell (123) · all languages

Topics: Mobile and desktop apps (615) · Cloud, DevOps and infrastructure (498) · Developer tools (405) · AI and machine learning (397) · Data and databases (310) · Web development (251) · Security and privacy (212) · Systems and embedded (189) · Games and graphics (148) · Documentation and education (78) · Science and research (60) · Finance and Web3 (54) · Robotics (8)

Why this exists

Finding a first issue is harder than it should be. Popular "good first issue" lists go stale, half the issues are already taken, and nobody tells you that a project requires a CLA, bans AI-generated code, or auto-closes pull requests from new accounts until after you have done the work.

Open Source Radar fixes the first part with data and the second part with a guide:

• Only issues you can actually pick up. Every listed issue is open, has no assignee, has no open or merged pull request linked to it, was updated in the last six months, and belongs to a project with recent commits.
• A directory of welcoming projects. Every active project with open newcomer issues, grouped by language, with stars and rules: projects directory.
• Organized the way you search. By language, by topic (web, developer tools, AI, data, DevOps, security, mobile, games, systems, docs, science, finance), by label level, and by project rules.
• Project rules up front. Each project is scanned for AI policies, Contributor License Agreements and DCO sign-off requirements, so you know what you are signing up for before you start.
• Always fresh. A GitHub Actions workflow rebuilds everything every 12 hours.
• A guide that covers the parts people skip. Not just "fork and clone": how to tell whether a project will review you, how to check an issue is really free, how to handle CLAs, DCO, AI policies and anti-spam bots, and what to do when your pull request is ignored or closed.

Start here

If you are...	Go to
Brand new to open source	Guide, chapter 1: why contribute and what counts
Ready to set up Git and GitHub	Chapter 2: set up your tools
Looking for a project	Projects directory and chapter 3: choose a project
Looking for an issue	The website or the issue index
About to open your first pull request	Chapter 5: your first pull request, step by step
Unsure about CLAs, DCO or AI rules	Chapter 6: rules to check before you start
Waiting on a review	Chapter 8: reviews, feedback and rejection
Setting up a project in a new language	Language quickstarts

The guide

1. Why contribute, and what counts
2. Set up your tools
3. Choose a project
4. Find an issue you can finish
5. Your first pull request, step by step
6. Rules to check before you start
7. Talking to maintainers
8. Reviews, feedback and rejection
9. Keep going

Plus a glossary and quickstarts for 14 languages.

How issues are selected

The data pipeline is a single Python script with no dependencies. On each run it:

1. Discovers projects through GitHub search: public, non-archived, non-fork repositories in 30 languages with at least 500 stars, commits in the last 60 days, and at least one open issue labeled good first issue or help wanted.
2. Collects issues carrying common newcomer labels (good first issue, help wanted, beginner, easy, E-easy, up-for-grabs and variants) through the GraphQL API.
3. Drops issues that are not available: assigned, locked, linked to an open or merged pull request, or not updated in 180 days.
4. Scans contribution files (CONTRIBUTING.md, AI_POLICY.md, AGENTS.md, pull request templates) for AI policies, CLA and DCO requirements, weekly.
5. Renders the Markdown pages in issues/ and the JSON behind the website.

All thresholds and label lists live in scripts/config.toml. The badges are detected automatically and can be wrong, so always read the project's own contributing guide. If you spot a mistake, open an issue.

Run it yourself

git clone https://github.com/TanbirRamim/open-source-radar.git
cd open-source-radar
export GITHUB_TOKEN=$(gh auth token)                  # any token with public read access
python3 scripts/radar.py all --languages "Rust,Go"    # Python 3.11+, no packages needed
python3 -m http.server -d site 8000                   # then open http://localhost:8000

Use the data

The full dataset behind the website is one JSON file, refreshed twice a day and free to use in your own bots, newsletters, dashboards or Discord servers:

curl -s https://tanbirramim.github.io/open-source-radar/data/issues.json | jq '.issues | length'

The format, with examples, is documented in docs/data.md.

Prefer a feed reader? Every language has an RSS feed of its 50 newest beginner issues, for example Python. All feeds are listed on the feed index.

Contributing

This project is itself a good first contribution. The good first issues are scoped so you can finish one in an evening, and every pull request gets a review. Ideas:

• Improve a guide chapter or a language quickstart, or add a new one.
• Add a label variant, language or topic keyword to scripts/config.toml.
• Report a project whose rules were detected incorrectly.
• Build something with the data and tell us about it.
• Improve the website or the pipeline, with a test in scripts/test_radar.py.

Read CONTRIBUTING.md first, and see the Code of Conduct. Maintainers who prefer their project not to be listed can open an issue and it will be excluded.

Contributors

Thank you to everyone who has made the radar better:

• @Nomee-123: the / shortcut that focuses the website search (#10)
• @yyashikasharmaa: the Robotics topic (#11)
• @dharma0009: RSS feeds for every language (#16)
• @PandaHUN777: Atom self links in the RSS feeds (#30)
• @nightcityblade: an end-to-end smoke test for the render pipeline (#34)
• @Shivani2965: a styled RSS feed index page (#36)
• @AkashGowdaNC: issue counts on the feed index, with empty feeds hidden, and RSS feeds per topic (#42, #52)
• @JiyaSinghal0604: the Hindi and Spanish translations of the guide (#44, #46)
• @nayan45633: lastBuildDate in the language feeds (#50)
• @Harshil-1603: feed item descriptions and categories, and the per-language RSS link on the website (#48, #49)
• @vamsikrishnaavasarala88-png: the Elixir quickstart (#51)
• @faresrafat3: RSS autodiscovery links on the feed page (#60)

Your name goes here with your first merged pull request. And if the radar helped you, please give it a star: it's the simplest way to help other newcomers find it.

Please, be a good citizen

Maintainers are people with limited time. When you use this list:

• Comment on or claim an issue only when you are really going to work on it.
• Do not open pull requests to many projects at once just to collect contributions.
• Follow each project's rules, including its policy on AI-generated contributions.
• If you stop working on something, say so, so someone else can pick it up.

If you found your first issue here, starring the repository helps other newcomers find it too.

License

MIT. Issue titles, descriptions and project metadata belong to their respective projects and are shown with links back to GitHub.