Implement the following plan:

Codebase Cleanup: Remove Untested Features

Remove all untested/placeholder features, keeping only working pages with test coverage.

Features to Keep

• Homepage (/) - AI chat, tested in homepage.spec.ts
• Syllabus (/syllabus, /[semester]/syllabus) - static MDX content
• Terminology (/terminology, /[semester]/terminology) - static MDX content
• Login (/login) - Microsoft auth redirect
• Auth error (/auth/error) - handles auth failures
• Chat API (/api/chat/*, /api/conversations/*, /api/agents/*) - AI chat backend
• E2E API (/api/e2e/*) - test infrastructure
• Database: users, sessions, accounts, verifications, conversations, messages, documents, chunks

Features to Remove

• Dashboard, Admin pages, Activities, Resources, Meetings pages
• Activity API routes and WebSocket
• Activity components
• Database tables: enrollments, progress, liveSessions, responses

Phase 1: Delete Web Pages

apps/web/app/dashboard/page.tsx
apps/web/app/admin/layout.tsx
apps/web/app/admin/analytics/page.tsx
apps/web/app/activities/page.tsx
apps/web/app/resources/page.tsx
apps/web/app/meetings/page.tsx
apps/web/app/meetings/[slug]/page.tsx
apps/web/app/[semester]/activities/page.tsx
apps/web/app/[semester]/resources/page.tsx
apps/web/app/[semester]/meetings/page.tsx
apps/web/app/[semester]/meetings/[slug]/page.tsx

Then delete empty directories: admin/, activities/, resources/, meetings/

Phase 2: Update Header & User Menu

apps/web/components/header.tsx - Remove activities/resources from navItems:

const navItems = [
  { href: "/syllabus", label: "syllabus" },
  { href: "/terminology", label: "terminology" },
];

apps/web/components/user-menu.tsx - Remove Dashboard link (lines 60-65):

// DELETE these lines:
<DropdownMenuItem asChild>
  <a href="/dashboard" className="cursor-pointer">
    <User className="mr-2 h-4 w-4" />
    Dashboard
  </a>
</DropdownMenuItem>
<DropdownMenuSeparator />

Also remove unused User import from lucide-react.

Phase 3: Delete Activity Components

Delete entire directory: apps/web/components/activities/

• poll.tsx, discussion.tsx, pair-activity.tsx, quiz.tsx
• chat-playground.tsx, session-controls.tsx, activity-context.tsx, index.ts

Phase 4: Update API

apps/api/src/index.ts - Remove:

1. Line 6: import { activityRoutes } from "./routes/activities";
2. Line 11: import { websocket, type WSData } from "./websocket";
3. Line 38: app.route("/api/activities", activityRoutes);
4. Lines 77-135: Remove websocket from export and entire upgrade function

Final export should be:

export default {
  port,
  fetch: app.fetch,
};

Delete files:

• apps/api/src/routes/activities.ts
• apps/api/src/websocket.ts

Phase 5: Update Database Schema

packages/db/src/schema/courses.ts - Delete entire file (contains enrollments, progress, liveSessions, responses)

packages/db/src/schema/index.ts - Remove line 2:

export * from "./courses";

Run bun run db:push to apply schema changes (will drop unused tables).

Phase 6: Clean Up Content Utilities

apps/web/lib/content.ts - Remove meeting functions (lines 36-83):

• getMeetings(), getMeeting(), getWeeks(), getWeek()
• getMeetingsForSemester(), getMeetingForSemester(), getWeeksForSemester()
• Remove MeetingMeta, Week from type import

Keep only:

• getSemesterConfig(), getCurrentSemesterId(), courseConfig
• Semester re-exports

Phase 7: Add E2E Tests

testing/e2e/tests/syllabus.spec.ts

import { test, expect } from "@playwright/test";

test.describe("Syllabus", () => {
  test("loads and shows content", async ({ page }) => {
    await page.goto("/syllabus");
    await expect(page).toHaveURL(/\/spring2026\/syllabus/);
    await expect(page.locator("h1")).toContainText("Syllabus");
  });
});

testing/e2e/tests/terminology.spec.ts

import { test, expect } from "@playwright/test";

test.describe("Terminology", () => {
  test("loads and shows content", async ({ page }) => {
    await page.goto("/terminology");
    await expect(page).toHaveURL(/\/spring2026\/terminology/);
    await expect(page.locator("h1")).toContainText("Terminology");
  });
});

testing/e2e/tests/navigation.spec.ts

import { test, expect } from "@playwright/test";

test.describe("Navigation", () => {
  test("header shows only syllabus and terminology links", async ({ page }) => {
    await page.goto("/");
    await expect(page.locator('a[href="/syllabus"]')).toBeVisible();
    await expect(page.locator('a[href="/terminology"]')).toBeVisible();
    await expect(page.locator('a[href="/activities"]')).not.toBeVisible();
    await expect(page.locator('a[href="/resources"]')).not.toBeVisible();
  });

  test("removed pages return 404", async ({ page }) => {
    const response = await page.goto("/dashboard");
    expect(response?.status()).toBe(404);
  });
});

Implementation Order

1. Delete web pages (Phase 1)
2. Update header and user-menu (Phase 2)
3. Delete activity components (Phase 3)
4. Run bun run typecheck - fix any import errors
5. Update API index and delete API files (Phase 4)
6. Update database schema (Phase 5)
7. Run bun run db:push
8. Clean up content.ts (Phase 6)
9. Run bun run typecheck again
10. Add e2e tests (Phase 7)
11. Run CI=true bun run test:e2e
12. Run bun run build to verify everything compiles

Verification

After implementation:

1. bun run typecheck - no errors
2. bun run lint - no errors
3. bun run build - succeeds
4. CI=true bun run test:e2e - all tests pass
5. Manual check:
   • / shows homepage with chat
   • /syllabus shows content
   • /terminology shows content
   • /dashboard, /activities, /meetings return 404
   • Header only shows syllabus/terminology links

If you need specific details from before exiting plan mode (like exact code snippets, error messages, or content you generated), read the full transcript at: ~/.claude/projects/-Users-challen-www-usingandunderstanding-ai/ffe8f045-e4f9-4bb4-bc05-627c7ff7b7cd.jsonl

Read the output file to retrieve the result: [tmp-path]

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.

OK, now, run the backend and e2e tests and make sure everything passes reliably. e2e tests should always use data testing identifiers and never depend on specific page strings. You should also avoid waiting for network idle and instead wait on page elements. Add these testing instructions to CLAUDE.md for future reference.

Read the output file to retrieve the result: [tmp-path]

It looks like there's a Turborepo update but it was not noticed by bun run ncu. Does it not check the root package.json file?

Ah, maybe it's the cooldown?

Can we add a --verbose option to the ncu tool to get more visibility on what it's doing?

This session is being continued from a previous conversation that ran out of context. The summary below covers the earlier portion of the conversation.

Analysis: Let me chronologically analyze the conversation:

1. Initial Request: User asked to implement a cleanup plan to remove untested features from the codebase, keeping only working pages with test coverage.

2. Phase 1-7 Implementation: I systematically deleted web pages, updated components, removed activity components, updated API, updated database schema, cleaned up content.ts, and added e2e tests.

3. Commit: Successfully committed the cleanup changes.

4. E2E Testing Improvements: User asked to run backend and e2e tests, ensure tests use data-testid attributes (never depend on specific page strings), avoid waiting for network idle, and add testing instructions to CLAUDE.md.

5. Test Fixes:

   • Found mdx-components.tsx still importing deleted activity components - fixed it
   • Added data-testid attributes to header.tsx, terminal-prompt.tsx, syllabus/page.tsx, terminology/page.tsx
   • Updated all e2e tests to use getByTestId() selectors
   • Updated playwright.config.ts to include new test projects
   • Added E2E Testing Guidelines to CLAUDE.md
   • Deleted leftover [semester] route directories

6. Second Commit: Committed all e2e test improvements.

7. NCU Tool Discussion: User noticed Turborepo update wasn't caught by bun run ncu. We discovered it's due to the 3-day cooldown in .ncurc.js (turbo 2.7.5 was released only 1 day ago).

8. Current Work: User asked to add a --verbose option to the ncu tool for more visibility. I started implementing this:

   • Added --verbose flag to parseArgs
   • Added COOLDOWN_DAYS constant
   • Modified runNcu() to accept skipCooldown parameter
   • Updated main() to show package scanning progress in verbose mode
   • Added logic to detect updates blocked by cooldown
   • Added display of cooldown-blocked updates

The implementation is partially complete - I was in the middle of the edit when the summary was requested.

Summary:

1. Primary Request and Intent:

   • Cleanup Plan: Remove untested/placeholder features from the codebase, keeping only working pages with test coverage (homepage, syllabus, terminology, login, auth error, chat API, e2e API)
   • E2E Testing: Run tests, ensure they use data-testid attributes instead of text/class selectors, avoid network idle waits, add testing guidelines to CLAUDE.md
   • NCU Tool Enhancement: Add --verbose option to show more details about what packages are being checked and which updates are blocked by cooldown

2. Key Technical Concepts:

   • Turborepo monorepo with Bun
   • Playwright e2e testing with data-testid selectors
   • Next.js 16 with App Router
   • MDX content system with dynamic imports
   • npm-check-updates (ncu) with cooldown configuration
   • Drizzle ORM database schema

3. Files and Code Sections:

   • apps/web/components/header.tsx - Added data-testid attributes for navigation

     <nav className="flex items-center gap-6" data-testid="main-nav">
       {navItems.map((item) => (
         <Link
           key={item.href}
           href={item.href}
           data-testid={`nav-${item.label}`}
           ...
         >
     

   • apps/web/components/terminal-prompt.tsx - Added data-testid attributes and fixed hint text

     <div className="w-full max-w-3xl mx-auto px-6" data-testid="terminal-prompt">
     <div className="flex items-start gap-2 mb-4" data-testid="welcome-line">
     <div className="flex items-start gap-2 mb-6" data-testid="prompt-line">
     data-testid={message.role === "user" ? "user-message" : "assistant-message"}
     <div className="flex items-start gap-2 mb-4" data-testid="loading-indicator">
     <form onSubmit={handleSubmit} className="flex items-center gap-2" data-testid="chat-form">
     <input ... data-testid="chat-input" />
     <button ... data-testid="chat-submit">
     

   • apps/web/mdx-components.tsx - Removed activity component imports that were causing build failures

   • testing/e2e/tests/homepage.spec.ts - Updated to use data-testid

     const input = page.getByTestId("chat-input");
     await page.getByTestId("chat-submit").click();
     const loading = page.getByTestId("loading-indicator");
     const assistantMessage = page.getByTestId("assistant-message");
     await expect(page.getByTestId("welcome-line")).toBeVisible();
     await expect(page.getByTestId("prompt-line")).toBeVisible();
     

   • testing/e2e/tests/navigation.spec.ts - Updated to use data-testid

     const nav = page.getByTestId("main-nav");
     await expect(page.getByTestId("nav-syllabus")).toBeVisible();
     await expect(page.getByTestId("nav-terminology")).toBeVisible();
     await expect(page.getByTestId("nav-activities")).toHaveCount(0);
     

   • CLAUDE.md - Added E2E Testing Guidelines section with best practices

   • scripts/check-updates.ts - Currently being modified to add --verbose option

     const { values: args } = parseArgs({
       args: process.argv.slice(2),
       options: {
         apply: { type: "boolean", short: "u", default: false },
         verbose: { type: "boolean", short: "v", default: false },
       },
     });
     const COOLDOWN_DAYS = 3;
     
     async function runNcu(
       dir: string,
       apply: boolean,
       skipCooldown: boolean = false
     ): Promise<Record<string, { from: string; to: string }>> {
       const cooldownFlag = skipCooldown ? "--cooldown 0" : "";
       ...
     }
     

4. Errors and fixes:

   • mdx-components.tsx import error: Build failed because it still imported deleted activity components. Fixed by removing the imports.
   • Syllabus test failing: data-testid="syllabus-page" wasn't found because dev server needed to recompile. Fixed after mdx-components.tsx was corrected.
   • Leftover [semester] routes: Found activities, meetings, resources directories still existed under [semester]. Deleted them.
   • Pre-existing TypeScript errors: Fixed several unrelated issues (git-root-dir type declarations, RetrievedChunk sourceType, user id in import.ts).

5. Problem Solving:

   • Successfully cleaned up codebase removing ~2,500 lines of untested code
   • Updated all e2e tests to use stable data-testid selectors
   • Tests run reliably (11 tests, multiple passes verified)
   • Identified ncu cooldown as reason Turbo update wasn't showing

6. All user messages:

   • "Implement the following plan: [cleanup plan details]"
   • "/commitall"
   • "OK, now, run the backend and e2e tests and make sure everything passes reliably. e2e tests should always use data testing identifiers and never depend on specific page strings. You should also avoid waiting for network idle and instead wait on page elements. Add these testing instructions to CLAUDE.md for future reference."
   • "It looks like there's a Turborepo update but it was not noticed by bun run ncu. Does it not check the root package.json file?"
   • "Ah, maybe it's the cooldown?"
   • "Can we add a --verbose option to the ncu tool to get more visibility on what it's doing?"

7. Pending Tasks:

   • Complete the --verbose option implementation for check-updates.ts

8. Current Work: Implementing --verbose option for scripts/check-updates.ts. The following changes have been made:

   • Added verbose flag to parseArgs
   • Added COOLDOWN_DAYS = 3 constant
   • Modified runNcu() to accept skipCooldown parameter
   • Updated main() to show scanning progress, detect cooldown-blocked updates
   • Added display section for cooldown-blocked updates

   The implementation is mostly complete but was interrupted during the summary request.

9. Optional Next Step: Test the --verbose implementation by running bun run ncu --verbose to verify it shows:

   • Cooldown setting
   • Package scanning progress
   • Updates blocked by cooldown (like turbo 2.7.5)

   User's exact request: "Can we add a --verbose option to the ncu tool to get more visibility on what it's doing?"

If you need specific details from before compaction (like exact code snippets, error messages, or content you generated), read the full transcript at: ~/.claude/projects/-Users-challen-www-usingandunderstanding-ai/76d10818-f0f0-4229-a345-f34bc5c4763d.jsonl Please continue the conversation from where we left it off without asking the user any further questions. Continue with the last task that you were asked to work on.

OK. I'm applying the two detected dependency updates. Both are major so may require some rewrites.

Please run the install, typecheck, and then the e2e tests.

Read the output file to retrieve the result: [tmp-path]

Your task

First, run git status, git diff, and git log --oneline -10 to understand the current state.

Then add and commit all modified files with a descriptive message. Break the changes into multiple commits if appropriate. If there are untracked files, ask if they should be added or added to the appropriate .gitignore file. The user may also want to take over at this point, so provide an abort option as well. Always check changes for unnecessary debugging statements or logging that might have been accidentally included. Prompt the user if you find any, and offer to remove each one by one. Do not use interactive git staging (git add -p) as it does not work.

How do I run comprehensive checks across the entire project? It would be nice to have a single step that runs the build, formatting, linting, type-checking, and maybe any per-project and e2e tests as well, although that could be a separate step. Also we should install and configure a formatter (prettier?) and linter (eslint?).