OpenAI Build Week · Education track · July 2026

Scaffolding for Taste

A polished homepage can hide the bad passes and judgment behind it. I wanted students to see what I kept, changed, and threw away, then reuse the same screenshot checks. After the cutoff, I used Codex with GPT-5.6 Sol to add the Human/AI route, Paper Constellation, guided project stories, reproduction briefs, and the tests behind them. The timeline keeps that new work separate from the site I already had.

12 post-cutoff checkpoints 7 new reproduction guides 1 retained GPT-5.6 Sol thread 0 runtime AI calls
Follow the evidence Play the guided tour Watch the 2:45 demo Open the live portfolio Read the design memory
Eligibility line July 13 · 9:00 AM PT

Before 09:00

Foundation

  • Research portfolio
  • COGS 125 framing
  • Living heuristics
  • Reciprocal 3D room

After 09:00

Extension

  • Human/AI profile
  • Guided evidence stories
  • Paper Constellation
  • Case studies + QA
The line marks the work judges should evaluate: the colored extension after July 13 at 9:00 AM. Everything gray was already here; this page asks judges to evaluate only the extension.
Why

I wanted visitors and students to see which design decisions survived review and reuse the checks behind them.

What

The extension added new reading routes, a publication map, guided case studies, reproduction guides, and responsive checks.

How

Codex accelerated bounded implementation passes; I reviewed fixed screenshots and decided what to keep, revise, or remove.

What changed after the line?

The timeline below links each public change to its commit. Read left to right by date; the three lanes separate public pages, evidence, and reusable notes.

Selected evidence · date on the horizontal axis

The foundation existed before the cutoff.

The orange rule fixes the eligibility boundary. Equal-sized nodes mark linked checkpoints; horizontal position encodes date.

Equal-sized checkpoint · lane = artifact role · x-position = authored date

  1. 01 · Draw the boundary

    Show what was already there.

    The portfolio, COGS 125 framing, living heuristics, first 2D/3D desk, activity workbench, and reciprocal cliff-room topology all predate 9:00 AM. The last named baseline, a765c5291, was committed before the window opened.

  2. 02 · Make the record readable

    Give the same research record two reading routes.

    d3f13be35 added the source-linked Human/AI research profile, publication catalog, and raw text alternatives. The canonical bibliography owns the facts; the new routes reduce visual decoding work.

  3. 03 · Guide the interpretation

    Explain commits, changed lines, and tokens separately.

    6b4b7bd59 added the scroll-led Build Rhythm story. f05f02b41 added the first case-study and reproduction layer, while b183e4b75 wrote the surviving constraints into design memory. The original workbench began on July 11; the guided teaching layer is the Build Week extension.

  4. 04 · Change the representation

    Make a desktop atlas and a mobile trail.

    855f1bce8 fixed the privacy-safe data contract; eeb0a5764 built Paper Constellation; and 6832a6a05 replaced the compressed mobile map with a chronological trail. I reviewed every relationship and coordinate; GPT-5.6 did not discover the research lineage.

  5. 05 · Show the work

    Give six earlier experiments and one new paper map their own case studies.

    Across f05f02b41, 1ec5d5f4f, 2070991b0, and the constellation work, six pre-existing interactions gained project pages alongside the new Paper Constellation. All seven include evidence and downloadable reuse guides.

  6. 06 · Publish the proof

    Test more than the best-looking screenshot.

    739d2ba28 published paired desk evidence under a fixed capture rubric. 1ae1ead39 and c52de6932 closed responsive and cross-engine gaps. The reciprocal room was the baseline; post-cutoff work polished its controls and evidence—it did not create the reciprocal room.

One record, two reading shapes

Paper Constellation is the clearest before-and-after within the Build Week extension. Both views use the same five papers, seven anonymous future nodes, three threads, and nine connections. Desktop makes comparison spatial; mobile puts chronology and readable titles first.

Desktop Paper Constellation atlas arranging accepted papers and anonymous future nodes across three research threads
Desktop atlas. Stable, reviewed coordinates support relationship tracing; position encodes thread membership.
Mobile Paper Constellation trail preserving three thread rails, accepted papers, and anonymous future nodes in chronological order
Mobile trail. The same record becomes a chronological path with readable labels and visible thread rails.

Seven experiments got their own case studies

Each page now answers four practical questions: why did I make this, what changed, what evidence survives, and what could someone else reuse?

How I used Codex during the week

I gave Codex one bounded change at a time, inspected the rendered result, and decided whether it clarified the page. GPT-5.6 Sol made those build-and-review passes faster; it did not make the research or acceptance decisions.

  1. 01Name the question

    Say what a rushed visitor should understand.

  2. 02Build one change

    Use Codex with GPT-5.6 Sol for a bounded implementation pass.

  3. 03Check the page

    Compare fixed routes, viewports, themes, and states.

  4. 04Decide

    Keep, revise, or revert based on meaning and access.

  5. 05Save the rule

    Write useful recurring critique into design memory.

Codex + GPT-5.6 accelerated

The implementation work

  • Tracing routes and implementation history
  • Building bounded visual and responsive passes
  • Writing focused source and browser checks
  • Comparing failures across fixed states

Human critique decided

What to keep

  • The research claim and eligibility boundary
  • Which visual relationships were defensible
  • Which directions to simplify or reject
  • Which lesson belonged in shared design memory

What another student can reuse

The checklist turns recurring screenshot critiques into checks for hierarchy, mobile layout, accessibility, evidence, and provenance. Students can adapt the checks, then decide which ones improve their own page.

Living design memory

Website design heuristics

A living markdown checklist for critiquing portfolio pages with clarity, restraint, and reflective practice.

Download Source 
Loading heuristics...

Exact evidence ledger and Codex provenance

Primary retained build thread: 019f652f-7154-7822-ad1c-daa5a066134b. The retained thread ran GPT-5.6 Sol at ultra reasoning effort. The model label records provenance, not isolated causality; model, prompt, retained context, implementation history, and human critique all changed together.

  1. a765c5291Named baseline: reciprocal cliff-room scene committed before the 9:00 AM eligibility line.
  2. d3f13be35Added the source-linked Human/AI profile, publication catalog, and raw alternatives.
  3. b183 · 6b4 · f05Recorded durable rules, added the Build Rhythm teaching story, and published the first experiment case studies and reproduction briefs.
  4. 855 · eebFixed the privacy-safe Paper Constellation contract and built the desktop atlas.
  5. 6832a6a05Recomposed the constellation as a chronological mobile trail.
  6. 739 · 1ec · 207Published paired desk evidence and completed provenance-rich experiment stories.
  7. 1ae · c52Closed responsive storytelling gaps and stabilized cross-engine visual checkpoints.

What comes next

Next I want to turn this into a classroom critique exercise with fixed viewports, one shared brief, comparable screenshots, and a worksheet for deciding which observations are worth keeping. Any claim about learning waits for an actual evaluation.

Credits

This page documents work completed for OpenAI Build Week with Codex and GPT-5.6. The site is a customized fork of the open-source al-folio academic theme and uses site-native graphics with static repository evidence.