Assessment and the evidence chain

Grade structure (current intent)

• 55% completion bucket: quizzes, weekly check-ins, and two design projects (with formative feedback).
• 45% capstone: milestone-based; groups up to 5; includes both group and individual reports.

Required GenAI use (with logging)

Generative AI use is required and must be documented. The course emphasizes design quality and evidence of iteration, not performative output.

See Using GenAI in this course without Goodharting it for the concrete protocol and evidence expectations.

Evidence chain requirements

• Signed git commits: all submissions are made via signed commits in the department GitHub Enterprise environment.
• Course GitHub organization: https://github.students.cs.ubc.ca/CPSC416-2025W-T2
• AI interaction log: maintain a running log (prompts, key excerpts, decisions taken, and what changed as a result).
• Traceability: the AI log entries should reference commits/PRs and the artifacts they influenced.
• Oral exam hooks (required structure): maintain a small, stable index of your reasoning in oral-exam-hooks/hooks.md (claims, failure scenarios, tradeoff, and one trace).

Specific formatting and templates will be provided in the relevant project/capstone assignments.