Criteria & Scoring

Each category is scored out of 10. Total: 60 points.

TL;DR: 6 categories, 10 points each

• Documentation: Comprehensive README, architecture overview, API docs, decision log, known issues
• AI Scan: Automated checks for security, transaction integrity, code quality, architecture, test coverage
• Code Quality: Clean, consistent, readable code with sensible structure and error handling
• DevOps: Docker Compose, meaningful commits, CI/CD, environment config, live deployment
• Testing: Core business logic coverage, edge cases, input validation, app stability
• Product: Working end-to-end functionality, usable UX, engaging demo, confident explanations

---

1. Documentation — 10 pts

Requirement	Notes
README with working setup instructions	Must get someone from clone to running in under 5 minutes
Architecture overview	Components, data flow, and rationale for key choices
API documentation	Endpoints, request/response format
Decision log	What was descoped, trade-offs made, key technical choices and why
Known issues	Honest account of what’s broken and what you’d do with more time

The decision log matters

Teams often skip the decision log. It’s worth points and demonstrates engineering maturity. An honest assessment of your technical choices and trade-offs is exactly what judges want to see.

---

2. AI Agent Scan Results — 10 pts

Scored from the automated scan output compiled at code freeze. Scan results are shared with judges before demos begin.

Critical issues are weighted heavily:

• Security vulnerabilities (hardcoded secrets, injection, missing validation)
• Transaction integrity failures (race conditions, data consistency issues)

Also scanned:

• Code quality (linting, dead code, complexity)
• Architecture (file structure, separation of concerns, circular dependencies)
• Test coverage (presence, quality, critical path coverage)
• Documentation completeness
• Dependency health (outdated packages, known vulnerabilities)

See the AI Agent Scan page for the full list of what gets checked.

---

3. Code Quality & Architecture — 10 pts

Requirement	Notes
Clean, consistent, readable code	Consistent style, sensible naming, no dead code
Sensible project structure	Separation of concerns, logical file organisation
Core business logic isolation	Logic is well-isolated and independently testable
Database schema design	Well-normalised, appropriate indexes, sensible relationships
Error handling and logging	Errors are caught, logged, and surfaced gracefully

---

4. DevOps & Deployment — 10 pts

Requirement	Notes
Meaningful commit history	Commits throughout the day, not one push at 16:55
Docker Compose works first try	docker compose up brings the whole stack up
CI/CD pipeline	Even a basic GitHub Actions workflow counts
Environment config	.env.example present, no hardcoded secrets
App is live and accessible	Live URL in README, accessible during demo

---

5. Testing & Reliability — 10 pts

Requirement	Notes
Core business logic test coverage	Critical paths, transaction integrity, edge cases
Edge cases handled	Invalid states, concurrent operations, error conditions
App stability	Doesn’t crash under normal usage
Input validation	Throughout the app, including core business logic
Graceful error responses	API returns sensible error messages and status codes

---

6. Product & Demo — 10 pts

Requirement	Notes
End-to-end functionality	The app actually works — full user journey from start to finish
Usable UX	Someone unfamiliar can navigate and use it
Clear and engaging demo	5 minutes, shows the full flow
Explains decisions confidently	Team can articulate architectural choices and trade-offs
Bonus: live audience interaction	Let audience members interact with your app during the demo

---

Scoring Summary

Category	Points
Documentation	/10
AI Agent Scan Results	/10
Code Quality & Architecture	/10
DevOps & Deployment	/10
Testing & Reliability	/10
Product & Demo	/10
Total	/60