Coding Education Platform

Part of the worlds-biggest-software-project initiative.

An open, AI-native coding education platform combining browser-based IDEs, automated grading, and Socratic mentorship for learners and institutions.

The Coding Education Platform delivers an integrated, browser-based environment for teaching, practising, and assessing programming. It targets higher-education faculty, bootcamps, K-12 CS programmes, and self-directed learners who need autograded assignments, sandboxed code execution, and meaningful feedback without paying institutional SaaS premiums or stitching together free tools.

Why Coding Education Platform?

Institutional autograding suites such as Codio and CodeGrade are priced at $50–200k per institution per year, locking out individual instructors and smaller bootcamps.
Codecademy and similar consumer products grade only correctness against fixed outputs; they cannot assess code quality, style, or efficiency, which are the skills that matter in professional practice.
Replit has a strong AI-embedded IDE but weak LMS integration, no psychometric analytics, and public-by-default student data — unsuitable for formal courses bound by FERPA or GDPR.
GitHub Classroom is free but ships no curriculum, rudimentary analytics, and a Git learning curve that creates significant instructor and student burden.
Existing autograders return pass/fail output with no explanation; LLMs now make line-level, learner-appropriate feedback technically and economically feasible at scale.

Key Features

Browser-Based Coding Environment

In-browser code editor supporting 20+ programming languages with no local installation
Isolated sandbox execution with configurable CPU, memory, and wall-clock limits per submission
Submission history and version tracking per learner
Optional Linux container per learner to eliminate environment inconsistency
Real-time collaborative coding for pair-programming and group assignments

Automated Grading and Assessment

Automated test case execution with per-test pass/fail, runtime, and memory output
Assignment distribution and submission collection with deadline enforcement
Rubric-based manual grading overlay alongside automated test results
Plagiarism detection within and across course sections
Bulk grade book export for instructors

LMS and Workflow Integration

LTI 1.3 grade passback to Canvas, Blackboard, Moodle, and D2L Brightspace
xAPI activity reporting to Learning Record Stores
Git-native submission workflow that teaches version control as a concurrent skill
SSO via OAuth 2.0 / OpenID Connect and SAML 2.0 for institutional deployments
REST API for assignment automation and roster management

AI-Augmented Feedback and Tutoring

Line-level LLM explanations describing precisely why a test case fails
Socratic tutoring mode that asks guiding questions instead of supplying solutions
AI-generated problem variants at configurable difficulty and topic to deter solution reuse
Code quality review covering naming, documentation, efficiency, and style adherence
Persistent learner weakness profiling across sessions with targeted remediation prescriptions

AI-Native Advantage

Traditional autograders stop at pass/fail. An LLM tutor can pinpoint the exact line that fails a test, adapt the explanation to the learner's demonstrated skill level, and steer them toward the bug through Socratic questions rather than giving away the answer. The same models can generate fresh problem variants on demand to undermine answer-copying from static banks, evaluate code quality dimensions that conventional unit tests ignore, and act as an always-available mentor for self-study learners who would otherwise lack one.

Tech Stack & Deployment

The platform is designed for both self-hosted and managed-cloud deployment. Sandbox execution can be built on a Judge0-compatible service running as an isolated microservice (avoiding GPL-3.0 entanglement of the host application) or on a custom POSIX/Linux-based container runtime. Standards alignment includes IMS LTI 1.3 for LMS launch and grade passback, IMS QTI v3.0 for assessment items, xAPI for activity tracking, and WCAG 2.2 for accessibility. Authentication uses OAuth 2.0 / OpenID Connect; institutional SSO uses SAML 2.0.

Market Context

The coding bootcamp market alone was valued at roughly $3.8–4.1 billion in 2025 and is projected to reach $6–7 billion by 2029–2031, with online delivery representing about 64% of share. Institutional autograding contracts run $50–200k per institution per year and K-12 CS curricula run $20–100 per student per year, while consumer subscriptions sit at $15–50 per month. Primary buyers include CS faculty, bootcamp operators, K-12 district curriculum coordinators, corporate L&D teams, and technical recruiters.

Project Status

This project is in the research and specification phase.
Contributions, feedback, and domain expertise are welcome.

Contributing

We welcome contributions from developers, domain experts, and potential users. See CONTRIBUTING.md for guidelines.

Important: All contributions must be your own original work or clearly attributed open-source material with a compatible licence. Copyright infringement and licence violations will not be tolerated and will result in immediate removal of the offending contribution. If you are unsure whether a piece of code, text, or other material is safe to contribute, open an issue and ask before submitting.

Licence

Licence to be determined. See discussion for context.