Where web hackathonsgetjudged fairly.

Was the development genuine?

Analyzes commits, branches, contributors, timestamps, pull requests, and repository history to verify genuine development activity and identify suspicious patterns such as post-deadline commits or minimal contribution from team members.

Is the deployment real?

Verifies that the submitted web application is actually deployed and that the claimed infrastructure, services, and deployment configuration are genuine and match the repository state from the end of the hackathon.

Do the features actually work?

Autonomously operates the website to verify required features, bonus features, user flows, and functional behavior, while collecting evidence such as screenshots and error logs to support its findings.

Is the engineering sound?

Evaluates overall engineering quality through specialized checks.

• Frontend: frontend implementation and structure.
• Backend: API and server-side implementation.
• Database: schema, queries, relationships, indexes, and migrations.
• Static Analysis: integration with industry-standard tools such as SonarQube, ESLint, Pylint, JUnit, pytest, and Jest, and metrics such as cyclomatic complexity, test coverage, code duplication, and code smells.

Is the app safe?

Identifies vulnerabilities, exposed secrets, insecure authentication and authorization, unsafe APIs, injection risks, vulnerable dependencies, and other web-security issues.

How does it behave under load?

Measures latency, response time, throughput, resource utilization, and application behavior under controlled load to assess scalability and reliability.

Does it solve the problem?

Determines whether the implemented solution actually addresses the hackathon's problem statement, objectives, target users, and judging criteria.

Can it be built and run again?

Verifies whether the submitted project can be built and run successfully from the provided repository and configuration.

Was this copied from another team?

Compares repositories from different teams to detect potentially shared or copied implementations.

Was the judging consistent?

Evaluates the judging process itself by detecting score outliers, excessive disagreement between agents, inconsistent treatment of similar projects, potential systematic bias, and cases requiring re-evaluation or human review.