[songoao25]

Courses exposed through the Task Center need a workflow separate from the chapter queue. This draft adds sequential group traversal and platform-state verification for videos, document reading, chapter synchronization, homework, topic discussions and AI practice. Unsupported thinking questions stay explicit, and Task Center failures remain isolated from chapter study.

The integration includes a discussion-board CLI, account-local configuration/cookies/cache, a setup wizard, question/answer parsing and durable review records. Substantive generated text is shown and recorded before requests; recording or validation failures block the affected submission. Cancelled, failed, unconfirmed and platform-accepted results remain distinct. Objective answers stay in runtime traces. Real duration requirements, request pacing and default-disabled quiz redo from #626 are preserved.

Scope and dependencies

This draft is the complete runnable integration. It includes prerequisites separately proposed in #637 (privacy/publication guards) and #638 (installable packaging); those smaller PRs are independently reviewable and should be considered first. Their overlap will need removal/rebasing after they merge. The remaining workflow modules share the CLI and local-data layer.

The branch starts from upstream main without importing the fork's development history. It preserves upstream app.py, license/author attribution, existing issue templates and compatibility exports. Public documentation follows the existing README; private handoffs, captures, account data and working notes are excluded.

Validation and limits

• Python 3.13 and 3.14: 366 offline tests passed, including current upstream regressions. Compile/publication checks and human-text audit self-tests passed.
• All new contribution commits were scanned with Gitleaks: no findings. Synthetic regressions cover credentials in logs/exceptions/reviews, cancellation and curl configuration injection.
• Independent review checked the current workflow/terminal/streaming changes; no new blocker was found. A local PTY verified pause, input restoration, resume and quit.
• Wheel/sdist builds passed. Linux installed-wheel and offline CI passed for the quality refactor; the latest commit (50ac130) passed Linux publication checks, all 366 offline tests and installed-wheel checks. The same commit passed Windows executable packaging, its help smoke test and artifact upload.
• No fresh live-account acceptance is claimed. Platform behavior can change, document-duration results require state readback and thinking questions remain unsupported.
• Text checks are conservative rules, not a complete semantic truthfulness guarantee. Optional model services receive questions/context, and local logs/reviews may contain course content; these flows are described in the README.

Upstream Codacy now passes at 50ac130 with zero added issues. CodeRabbit reports success for this draft, but draft status skips its full technical review.

The quality refactor addresses the upstream Codacy findings through documentation normalization, smaller functions, removal of unused code and targeted security fixes. It does not broadly disable the upstream analysis rules.

Maintainer feedback on scope and API boundaries is welcome before promoting this draft for merge.

[coderabbitai]

Important

Draft PR not reviewed

Draft PRs are not automatically reviewed by default.

• Trigger a manual review

To automatically review draft PRs, update your CodeRabbit configuration:

reviews:
  auto_review:
    drafts: true

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Autopilot is currently an internal CodeRabbit preview.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}