Harden capture: Readability, item index, Reader updates, toast retry

.github/workflows/ci.yml

@@ -0,0 +1,36 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Setup PNPM
+        uses: pnpm/action-setup@v3
+        with:
+          version: 10
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: 20
+          cache: pnpm
+
+      - name: Install deps
+        run: pnpm install --frozen-lockfile
+
+      - name: Typecheck
+        run: pnpm typecheck
+
+      - name: Test
+        run: pnpm test
+
+      - name: Build
+        run: pnpm build

.github/workflows/release.yml

@@ -17,7 +17,7 @@ jobs:
       - name: Setup PNPM
         uses: pnpm/action-setup@v3
         with:
-          version: 9
+          version: 10
 
       - name: Setup Node
         uses: actions/setup-node@v4
@@ -28,6 +28,12 @@ jobs:
       - name: Install deps
         run: pnpm install --frozen-lockfile
 
+      - name: Typecheck
+        run: pnpm typecheck
+
+      - name: Test
+        run: pnpm test
+
       - name: Build
         run: pnpm run build
 

.gitignore

@@ -2,5 +2,4 @@ node_modules/
 dist/
 .DS_Store
 
-ui/
 

README.md

@@ -13,26 +13,32 @@
 - **Inspire new ideas.** Beyond summaries, the assistant can remix your saved content: propose blog outlines, connect disparate articles, prompt you with “What if…” questions—anything that nudges you to create rather than collect.
 
 ### Feature List
-- Capture the current tab and extract readable content with the built-in readability fallback.
+- Capture the current tab and extract readable content with bundled Mozilla Readability, falling back to `article`/`main` text.
 - Auto-tag new items by calling an OpenAI-compatible `chat/completions` endpoint that you configure at runtime—no hard-coded hosts, no forced vendors.
 - Optionally save new captures to Pinboard and/or Readwise Reader. Import Pinboard tags, and export selected items to GoodLinks or Readwise.
+- Readwise Reader creates or updates the existing document, including tags.
+- After save, a page toast can retry a failed destination and edit tags.
 - Store everything locally; no backend service required.
 - Navigate a full options dashboard for configuring LLM models, privacy modes, tag behaviour, and integrations.
 - Inspect the Options-page Data Preview before saving: it shows local storage, Chrome sync storage, configured outbound API destinations, privacy-mode payload fields, and screenshot-safe secret states.
 
 ### Getting Started
-- Install Node 18+.
-- `npm install` to pull dependencies.
-- `npm run build` for a one-off bundle (or `npm run watch` while hacking).
+- Install Node 18+ and [pnpm](https://pnpm.io/).
+- `pnpm install` to pull dependencies.
+- `pnpm test` and `pnpm typecheck` for the current automated checks.
+- `pnpm build` for a one-off bundle (or `pnpm watch` while hacking).
 - Load the unpacked extension from the freshly created `dist/` directory in Chrome.
 - Hop into the Options page to plug in your LLM base URL, model name, and optional API key. Paste Pinboard and Readwise tokens only if you want those destinations, then use the capture checkboxes to enable or skip them on save.
+- Page extraction uses the bundled Mozilla Readability library, then `article`/`main` text, then `document.body.innerText`.
 
 ### Packaging & Release Ritual
 - Bump version in `package.json` (the build script mirrors it into `dist/manifest.json` automatically).
 - `npm run build` then `npm run package` (see `scripts/package.mjs`) to produce `tldr-v<version>.zip`.
 - Upload to the Chrome Web Store with updated screenshots and data-collection notes. Use the Options-page Data Preview as the product-visible source for privacy fields: it distinguishes local data, synced settings, user-configured remote API calls, and the absence of remotely hosted executable code.
 - Tag the release (`git tag v<version> && git push origin --tags`), draft GitHub notes, and attach the zip for sideloaders.
 
+Page extraction bundles [Mozilla Readability](https://github.com/mozilla/readability) (Apache-2.0).
+
 ### Contributing & Feedback
 - I’m dogfooding the tool to tame my own reading queue, but the north star is a community-sourced thinking partner. Open issues for features you crave, integrations you rely on, or moments where the flow breaks.
 - Want to explore the “report” or “idea lab” concepts early? Reach out—I’d love to collaborate on prototypes.

docs/release/chrome-web-store-publishing-checklist.md

@@ -85,8 +85,8 @@ This checklist records the current Chrome Web Store publishing flow for `tldr`.
 - [ ] `activeTab`: accesses the page only when the user previews or saves the active tab.
 - [ ] `scripting`: injects the extraction/readability script into the active tab after user action.
 - [ ] `contextMenus`: opens settings from the extension action menu.
-- [ ] `notifications`: verify whether still needed; remove if unused.
-- [ ] `alarms`: verify whether still needed; remove if unused.
+- [x] `notifications`: removed; unused.
+- [x] `alarms`: removed; unused.
 - [ ] Host permissions:
   - Pinboard API for tag import and bookmark sync.
   - Readwise API for Reader export.

package.json

@@ -7,6 +7,7 @@
     "build": "node scripts/build.mjs",
     "watch": "node scripts/build.mjs --watch",
     "test": "vitest run",
+    "typecheck": "tsc --noEmit",
     "clean": "rm -rf dist",
     "package": "node scripts/package.mjs"
   },
@@ -15,5 +16,8 @@
     "esbuild": "^0.21.5",
     "typescript": "^5.5.4",
     "vitest": "^4.1.8"
+  },
+  "dependencies": {
+    "@mozilla/readability": "^0.6.0"
   }
 }

pnpm-workspace.yaml

@@ -1,2 +1,5 @@
+packages:
+  - '.'
 allowBuilds:
   esbuild: true
+

scripts/build.mjs

@@ -4,38 +4,42 @@ import { existsSync } from 'node:fs';
 
 const watch = process.argv.includes('--watch');
 
-async function main() {
-  if (existsSync('dist')) await rm('dist', { recursive: true, force: true });
-  await mkdir('dist', { recursive: true });
+const shared = {
+  bundle: true,
+  sourcemap: true,
+  target: ['chrome114'],
+  platform: 'browser',
+  logLevel: 'info'
+};
 
-  const opts = {
-    entryPoints: [
-      'src/background/index.ts',
-      'src/ui/options.ts',
-      'src/content/extract.ts'
-    ],
-    outdir: 'dist',
-    outbase: 'src',
-    bundle: true,
-    sourcemap: true,
-    format: 'esm',
-    target: ['chrome114'],
-    platform: 'browser',
-    logLevel: 'info'
-  };
+const appBuild = {
+  ...shared,
+  entryPoints: [
+    'src/background/index.ts',
+    'src/ui/options.ts',
+    'src/content/extract.ts'
+  ],
+  outdir: 'dist',
+  outbase: 'src',
+  format: 'esm'
+};
 
-  if (watch) {
-    const ctx = await context(opts);
-    await ctx.watch();
-    console.log('Watching for changes…');
-  } else {
-    await build(opts);
+const extraBuilds = [
+  {
+    ...shared,
+    entryPoints: ['src/content/readability-global.ts'],
+    outfile: 'dist/readability.js',
+    format: 'iife'
+  },
+  {
+    ...shared,
+    entryPoints: ['src/content/toast.ts'],
+    outfile: 'dist/content/toast.js',
+    format: 'iife'
   }
+];
 
-  // Copy static files
-  await cp('static', 'dist', { recursive: true });
-
-  // Sync manifest version with package.json
+async function syncManifestVersion() {
   const pkg = JSON.parse(await readFile('package.json', 'utf-8'));
   const manifestPath = 'dist/manifest.json';
   try {
@@ -47,4 +51,27 @@ async function main() {
   }
 }
 
+async function copyStatic() {
+  await cp('static', 'dist', { recursive: true });
+  await syncManifestVersion();
+}
+
+async function main() {
+  if (existsSync('dist')) await rm('dist', { recursive: true, force: true });
+  await mkdir('dist', { recursive: true });
+  await copyStatic();
+
+  if (watch) {
+    for (const opts of [appBuild, ...extraBuilds]) {
+      const ctx = await context(opts);
+      await ctx.watch();
+    }
+    console.log('Watching for changes…');
+    return;
+  }
+
+  await build(appBuild);
+  for (const opts of extraBuilds) await build(opts);
+}
+
 main().catch((e) => { console.error(e); process.exit(1); });

src/background/exporters.test.ts

@@ -0,0 +1,97 @@
+import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest';
+import type { Settings } from '@common/types';
+
+vi.mock('@common/storage', () => ({
+  getSettings: vi.fn(),
+  getSecret: vi.fn()
+}));
+
+vi.mock('./readwise', () => ({
+  syncReadwiseDocument: vi.fn()
+}));
+
+import { getSecret, getSettings } from '@common/storage';
+import { syncReadwiseDocument } from './readwise';
+import { exportToGoodlinks, exportToReadwise } from './exporters';
+
+const settings: Settings = {
+  llm: { baseUrl: 'https://llm.example.test/v1', model: 'test', jsonMode: true, maxChars: 1000 },
+  pinboard: { shared: true, toread: false },
+  readwise: { apiTokenRef: 'readwise_token' },
+  tagging: { knownTagLimit: 200, dedupeThreshold: 82, aliases: {} },
+  privacy: { mode: 'title_excerpt' }
+};
+
+describe('exportToGoodlinks', () => {
+  let create: ReturnType<typeof vi.fn>;
+
+  beforeEach(() => {
+    vi.useFakeTimers();
+    create = vi.fn().mockResolvedValue(undefined);
+    (globalThis as unknown as { chrome: unknown }).chrome = {
+      tabs: { create }
+    };
+  });
+
+  afterEach(() => {
+    vi.useRealTimers();
+  });
+
+  it('opens a GoodLinks URL for each item', async () => {
+    const pending = exportToGoodlinks([
+      { url: 'https://example.test/a', title: 'A', tags: ['ai'] },
+      { url: 'https://example.test/b', title: '', tags: [] }
+    ]);
+    await vi.runAllTimersAsync();
+    await expect(pending).resolves.toBe(2);
+
+    expect(create).toHaveBeenCalledTimes(2);
+    expect(String(create.mock.calls[0]?.[0]?.url)).toContain('goodlinks://add?url=https%3A%2F%2Fexample.test%2Fa');
+    expect(String(create.mock.calls[0]?.[0]?.url)).toContain('tags=ai');
+  });
+
+  it('skips items that fail to open and still counts successes', async () => {
+    create
+      .mockRejectedValueOnce(new Error('blocked'))
+      .mockResolvedValueOnce(undefined);
+    const pending = exportToGoodlinks([
+      { url: 'https://example.test/a', title: 'A', tags: [] },
+      { url: 'https://example.test/b', title: 'B', tags: [] }
+    ]);
+    await vi.runAllTimersAsync();
+    await expect(pending).resolves.toBe(1);
+  });
+});
+
+describe('exportToReadwise', () => {
+  beforeEach(() => {
+    vi.useFakeTimers();
+    vi.mocked(getSettings).mockResolvedValue(settings);
+    vi.mocked(getSecret).mockResolvedValue('rw_token');
+    vi.mocked(syncReadwiseDocument).mockResolvedValue({ id: 'doc-1' });
+  });
+
+  afterEach(() => {
+    vi.useRealTimers();
+  });
+
+  it('throws when no token is configured', async () => {
+    vi.mocked(getSecret).mockResolvedValue(undefined);
+    await expect(exportToReadwise([{ url: 'https://example.test/a', title: 'A', tags: [] }])).rejects.toThrow(
+      'Readwise token not set'
+    );
+  });
+
+  it('exports items and continues after an individual failure', async () => {
+    vi.mocked(syncReadwiseDocument)
+      .mockRejectedValueOnce(new Error('nope'))
+      .mockResolvedValueOnce({ id: 'doc-2' });
+
+    const pending = exportToReadwise([
+      { url: 'https://example.test/a', title: 'A', tags: ['ai'] },
+      { url: 'https://example.test/b', title: 'B', tags: [] }
+    ]);
+    await vi.runAllTimersAsync();
+    await expect(pending).resolves.toBe(1);
+  });
+});

src/background/exporters.ts

@@ -1,5 +1,5 @@
 import { getSecret, getSettings } from '@common/storage';
-import { postReadwiseReaderDocument } from './readwise';
+import { syncReadwiseDocument } from './readwise';
 
 export type SimpleItem = { url: string; title: string; tags: string[] };
 
@@ -29,7 +29,7 @@ export async function exportToReadwise(items: SimpleItem[]): Promise<number> {
   let count = 0;
   for (const it of items) {
     try {
-      await postReadwiseReaderDocument(token, {
+      await syncReadwiseDocument(token, {
         url: it.url,
         title: it.title || undefined,
         tags: it.tags || [],