Sprint Competition Rules

Please read these rules carefully before entering a sprint. By submitting a project, you agree to abide by all rules outlined below.

1. Eligibility Criteria

To participate in a StartupTribunal sprint, you must meet the following requirements:

  • You must have a registered StartupTribunal account with a completed hacker profile.
  • You must be at least 18 years of age or the age of majority in your jurisdiction.
  • You may submit only one project per sprint campaign.
  • Team submissions are permitted, but only one team member should submit the project. All team members must have individual hacker profiles.
  • Employees and contractors of StartupTribunal and Genesis Protocol are not eligible to participate.

2. Submission Requirements

All submissions must meet the following criteria to be accepted:

  • Your project must be hosted in a public GitHub repository.
  • The repository must not be archived and must contain at least the campaign's min_commits_required commits during the sprint window. Sprint #1 requires 5 commits. Only commits inside the sprint window count toward this minimum — pre-existing work in your repo doesn't count, regardless of how much there is.
  • The repository size must not exceed 100 MB.
  • If you provided a GitHub username on your hacker profile, at least one of the in-window commits must be by that username. Teammates can contribute, but the submitter must show up in the history.
  • Your submission must include a project title, repository URL, description, and technology stack.
  • The repository URL is locked upon first save and cannot be changed after submission.
  • Submissions must be finalized before the campaign deadline. Draft submissions that are not submitted by the deadline will not be judged.
  • Your project must comply with our AI Usage Policy.

3. Judging Criteria

Submissions are evaluated by VibeJudge, our AI-powered judging system. Each project is scored across four dimensions with the following weights:

  • Innovation (60%) — Technical novelty, creative problem-solving, architecture elegance, README quality, and demo potential.
  • Bug Hunting (15%) — Code quality, security practices, test coverage, error handling, and dependency hygiene.
  • Performance (15%) — Architecture design, database design, API design, scalability considerations, and resource efficiency.
  • AI Detection (10%) — Commit authenticity, development velocity, authorship consistency, iteration depth, and AI generation indicators. Scoring is adjusted based on the campaign's AI policy mode.

Your final score is a weighted combination of these four dimensions. Scores are calculated after the submission deadline when the campaign enters the judging phase.

4. Prize Distribution

Prizes are distributed from the campaign's total prize pool, which grows from entry fees and sponsor contributions. The default distribution is:

  • 🥇 1st Place — Receives the largest share of the prize pool as defined by the campaign's distribution percentages.
  • 🥈 2nd Place — Receives the second-largest share of the prize pool.
  • 🥉 3rd Place — Receives the third-largest share of the prize pool.

Exact prize amounts and distribution percentages are displayed on the campaign page before the sprint begins. Finalists who place in the top tier but outside the top 3 receive recognition but no monetary prize.

5. Disqualification Conditions

Submissions may be disqualified at two stages of the judging process:

Pre-filter (before AI judging): Automated screening checks run before VibeJudge analysis. Your submission will be disqualified if any of the following are detected:

  • Empty repository (0 commits) or insufficient commit history (fewer than the campaign's min_commits_required during sprint window — Sprint #1 requires 5)
  • No commits by registered GitHub username (if you provided a GitHub username in your profile)
  • Suspicious files such as .gitmodules, symlinks, or archive files (.tar.gz, .zip, .rar, .7z, .war)
  • README prompt injection attempts targeting the AI judge
  • Repository exceeding the 100 MB size limit
  • Framework or boilerplate detection with no original work

Post-analysis (after AI judging): After VibeJudge scores your submission, the disqualification engine evaluates the results. Your submission may be disqualified if:

  • Innovation score is extremely low combined with framework or non-original work indicators
  • Git history analysis flags the repository as a framework with a low innovation score

Disqualified submissions are excluded from the leaderboard and are not eligible for prizes. Disqualification is a terminal state and cannot be reversed.

6. Entry Fee Policy

Each sprint requires a $10entry fee, payable via Flutterwave at the time of submission. The entry fee contributes directly to the campaign's prize pool.

Sponsored entries: Some campaigns offer sponsored spots where the entry fee is waived. Sponsored spots are limited and allocated on a first-come, first-served basis. If a sponsored spot is available, you will not be charged an entry fee.

Refunds: Entry fees are non-refundable once payment is confirmed, except in cases where the campaign is cancelled before judging begins.

Payment verification: If the payment webhook is delayed, the system will automatically verify your payment status. You will receive confirmation once your payment is processed.

7. Sponsor Survey

After the submission deadline, a time-limited sponsor survey may open. Completing the survey earns bonus points added to your final score. The survey window dates are announced with each campaign.

8. Code of Conduct

  • Treat all participants, organizers, and sponsors with respect.
  • Do not attempt to manipulate or game the judging system.
  • Do not submit projects that contain malicious code, hate speech, or illegal content.
  • Report any issues or concerns to the organizers promptly.

FAQ — The Cheat Code

Detailed answers to every question we get from serious builders. The above sections are the rules; this is the playbook for actually winning. Searchable, deep-linkable, and pulled directly from the gates that score your submission.

The Sprint Window Rule

The sprint window is the date range your submission must be built within. It is the most-misunderstood gate — read this carefully.

Pre-filter Gates (run before AI scoring)

Crown runs every submission through a deterministic pre-filter before sending it to VibeJudge. These checks happen in order; the first one that fails ends the analysis.

VibeJudge AI Scoring

If you pass pre-filter, your repo is cloned and analyzed by 5 AI agents on Amazon Bedrock. Each produces a score, a verdict, and file:line evidence.

Eligibility & How to Submit

Entry Fee & Prizes

AI Usage Policy (Cheat-code Edition)

StartupTribunal is built for vibe coders. AI tools are welcome — but understanding what triggers and what protects your AI Detection score is the difference between a good submission and a disqualification.

Disqualification — What Gets You Kicked Out

Disqualification is a terminal state — it cannot be reversed. There are two stages where it can happen.

After the Sprint Closes

Code of Conduct

9. Related Policies

For details on permitted AI tool usage and disclosure requirements, please review our AI Usage Policy.