Find your next coding project in under 30 seconds.
Get Started • Features • How It Works • Contribute • Docs • Discussions
Most beginner developers complete tutorials but struggle with the next step: what to build on their own.
DevPath was created to solve that problem.
Instead of endlessly searching for project ideas, users can enter their skills, interests, experience level, and available time to instantly get personalized project recommendations with:
- Step-by-step roadmaps
- Curated learning resources
- Beginner-friendly starter code
- Clear progression paths
The goal is simple:
Help developers move from “learning syntax” to “building real projects.”
| Feature | Description |
|---|---|
| Personalized recommendations | Matches projects based on skills, level, interest, and time |
| Rule-based scoring engine | Transparent and explainable recommendation logic |
| Beginner-friendly architecture | Small, modular, and easy to understand |
| Starter code templates | Downloadable boilerplate code for projects |
| Guided roadmaps | Structured learning and implementation flow |
| Open-source contribution ready | Designed for GSSoC and beginner contributors |
User inputs Scoring engine Output
----------- -------------- ------
Skills (Python, HTML) --> +3 per skill match --> Top 3 projects
Level (Beginner) --> +2 if level matches with:
Interest (Data) --> +2 if interest matches - Roadmap
Time (Low) --> +1 if time matches - Resources
- Starter code
The algorithm lives entirely in utils/recommender.py. Weights are named
constants at the top of the file — easy to tune without reading the whole module.
flowchart LR
A[User Input Form]
--> B[Flask Routes]
B --> C[Recommendation Engine]
C --> D[projects.json Dataset]
C --> E[Top 3 Project Matches]
E --> F[Project Roadmap]
E --> G[Starter Code Download]
E --> H[Learning Resources]
git clone https://github.com/komalharshita/devpath.git
cd devpath
python -m venv venv
source venv/bin/activate # Windows: venv\Scripts\activate
pip install -r requirements.txt
python app.pyhttp://127.0.0.1:5000 — that is the entire setup.
# Verify everything works
python tests/test_basic.py
# 27 passed, 0 failed out of 27 testsdevpath/
├── app.py Entry point (30 lines)
├── routes/main_routes.py 5 HTTP routes as a Blueprint
├── utils/
│ ├── data_loader.py Reads projects.json
│ ├── recommender.py Scores + filters projects
│ └── file_server.py Serves starter code safely
├── data/projects.json Project dataset (extend this)
├── templates/ Jinja2 HTML
├── static/ CSS + vanilla JS
├── starter_code/ 7 starter templates
├── tests/test_basic.py 27 tests
└── docs/ Architecture + contribution guides
| Method | Path | Returns |
|---|---|---|
| GET | / |
Homepage with skill form |
| POST | /api/recommend |
JSON — top 3 matched projects |
| GET | /project/<id> |
HTML — full project detail page |
| GET | /project/<id>/code |
JSON — starter code content |
| GET | /project/<id>/download |
File — starter code download |
The dataset is a plain JSON file. Add a new project by appending to
data/projects.json:
{
"id": 8,
"title": "Todo CLI App",
"skills": ["Python"],
"level": "Beginner",
"interest": "Automation",
"time": "Low",
"description": "A command-line task manager that saves to JSON.",
"features": ["Add/remove tasks", "Mark complete", "Filter by status"],
"tech_stack": ["Python", "json module"],
"roadmap": ["Step 1: Define data structure", "Step 2: Write add_task()"],
"resources": ["Python docs: https://docs.python.org"],
"starter_code": "starter_code/todo_cli.py"
}No backend changes needed. The engine picks it up on the next request.
New to open source? This project is designed for beginners.
Start here:
- Look for issues labeled
good first issue - Read
CONTRIBUTING.md - Ask questions in Discussions
- Follow the PR template
- Start with documentation or UI improvements
Maintainers review beginner PRs with guidance and feedback.
This project is designed to be contributed to. The codebase is small, modular, and thoroughly documented. If this is your first open-source contribution, this is a good place to start.
Step-by-step process:
1. Browse issues github.com/komalharshita/devpath/issues
2. Comment to claim Leave a comment before starting
3. Fork the repo Fork button on GitHub
4. Create a branch git checkout -b feat/your-change
5. Make the change Follow the code style in CONTRIBUTING.md
6. Run tests python tests/test_basic.py
7. Push and open a PR Use the template in CONTRIBUTING.md
Branch naming:
feat/description New feature
fix/description Bug fix
docs/description Documentation only
data/description New projects in the dataset
style/description CSS or visual changes
test/description Test additions or fixes
DevPath is an active GSSoC 2026 project. Contributions are mentored and welcomed at all skill levels.
Before starting work: comment on the issue, then fork and branch. Do not open a PR for an issue that is not assigned or claimed in the comments.
Read the full onboarding guide: docs/contribution_guide.md
Have questions, ideas, or feature suggestions?
- Open a discussion: https://github.com/komalharshita/devpath/discussions
- Report bugs through Issues
- Suggest new project ideas
- Help improve beginner onboarding
Community contributions and feedback are always welcome.
Thanks to all the amazing people who contribute to DevPath.
| Document | Description |
|---|---|
| README.md | This file — setup, structure, contributing |
| CONTRIBUTING.md | Code style, branch naming, PR template |
| CODE_OF_CONDUCT.md | Community standards |
| docs/architecture.md | System design and data flow |
| docs/contribution_guide.md | Beginner onboarding (8 steps) |
| docs/project_overview.md | What DevPath is and why |
| docs/github_issues.md | All 12 issues with full descriptions |
All contributors are expected to follow the Contributor Covenant Code of Conduct.
MIT — see LICENSE
DevPath — open source, built for learners, by learners.