[renrenmimi]

Summary

This adds CHANGELOG.md at the repository root. It records this audit round in English, which is the language of the project's development docs. Each change links its PR, #9–#23, and the file closes with the things deliberately left as they are.

Changes

• CHANGELOG.md, new. The entries are grouped into bugs, teaching content, interaction and accessibility, layout, performance and project. A "Not changed" section closes the file, and a PR link list follows it.

Verification

• Every entry was checked against its PR description and against the source on master (f7adc58).
• Comparison method: the before and after figures were measured on local production builds of ae5feb8 (before) and f7adc58 (after). Each build ran with next start and was measured in headless Chrome 155 through playwright-core 1.62.1, on macOS. Every measurement used a fresh context with an empty cache and no throttling.
• Measured results:
  • Prefetch, 390×844 phone, English: after the load event, /, /types and /mindset made 30–33 requests (378–415 KB) in seven seconds. They now make 0.
  • Fonts, English /types: 10 files (362 KB) before, 3 files (86 KB) now.
  • Idle styling, desktop: 23 s after load, /, /types, /generics and /type-magic did 180–181 style recalculations in 3 s. They now do 4–7.
  • Text contrast, 12 pages: every text element was checked against WCAG AA. 1,451 elements failed in the light theme and 1,360 in the dark theme; now none fail. Two kinds of false positive are left out of these counts: unticked checkboxes, whose ✓ is transparent on purpose, and buttons on a gradient.
  • Clipping, 12 pages, English and Chinese: at 360 and 390 px, 12–14 elements ran past the viewport, and 1 did at 1440 px. Now none do.
• The hydration figure, about one cold load in seventeen, is the audit's production measurement. Locally both builds hydrated cleanly in 120 cold starts, so the race does not reproduce locally and no local before/after is claimed for it.

🤖 Generated with Claude Code

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
tser	Ready	Preview	Oct 10, 2026 6:40am UTC

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-10T06:41:20.166610Z	61f3215	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[vercel]

Deployment failed for project tser with the following error:

Resource is limited - try again in 24 hours (more than 100, code: "api-deployments-free-per-day").

Learn More: https://vercel.com/renrenmimis-projects?upgradeToPro=build-rate-limit