Course Audit

Students begin descriptive writing about real websites in Week 1, translate description into Markdown blueprints in Week 2, and have a deployed personal landing page on their subdomain by Week 4. Hardware foundations (M.CS.1B → H.CS.1A), tokenization (M.DA.1A → H.DA.1A), and decomposition (M.AP.1B → H.AP.1B) surface as threads in service of the build, not as standalone weeks before building can begin. By Week 6, every student has a portfolio of tested Logic Blueprints AND a deployed landing page that proves them.

Students learn to verify AI output, detect confabulation and fabrication, map algorithmic bias, and understand why local AI protects community data - while continuing to iterate on their landing page from Act 1. The audit work runs alongside the build, using students' own AI interactions as primary evidence rather than curated examples. By Week 12, every student has an AI Audit Report with documented evidence and a scoped concept for their Act 3 PAR tool.

Students transition from their personal landing page (built in Act 1, refined through Act 2) to a cross-curricular PAR tool that serves another class or community need. The Prompt-to-App workflow has been in use since Week 4; Act 3 applies it to a more ambitious target with multiple components, real users, and design decisions that matter. The result is a deployed tool with a recorded Architectural Defense, alongside the student's complete portfolio.