feat: Lynote Notes open-source demo space (landing page + ready Gradio app)

README.md

@@ -1,10 +1,90 @@
 ---
 title: Lynote Notes
-emoji: 📈
-colorFrom: pink
-colorTo: gray
+emoji: 📝
+colorFrom: indigo
+colorTo: green
 sdk: static
-pinned: false
+pinned: true
+license: mit
+short_description: "Source-grounded AI notes with citations (open source)."
+tags:
+  - ai-notes
+  - note-taker
+  - source-grounded
+  - citations
+  - rag
+  - local-first
+  - bilingual
+  - open-source
+  - python
+  - cli
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 📝 Lynote Notes
+
+**Source-grounded, bilingual, local-first AI note taker.** Turn documents, web pages, recordings, videos and YouTube links into structured notes where every bullet cites the exact source chunk it came from — so you can verify, edit and reuse notes instead of trusting a black box.
+
+**Open source (MIT):** [github.com/lynote-ai/lynote-notes](https://github.com/lynote-ai/lynote-notes)
+
+## The problem
+
+AI note tools usually give you a summary you cannot check. If a note says "BM25 is used for retrieval", you should be able to click through to the sentence that said it. Lynote Notes makes traceability the default contract: notes are generated from your sources, and every claim links back to a chunk — including a timestamp for audio and video.
+
+## Quickstart
+
+```bash
+git clone https://github.com/lynote-ai/lynote-notes.git
+cd lynote-notes
+pip install -e .
+
+lynote-notes add lecture.md
+lynote-notes add https://example.com/post
+lynote-notes note --title "Weekly reading"
+lynote-notes ask "What did we decide about pricing?"
+lynote-notes export --note note_xxx --format anki --out cards.tsv
+```
+
+Optional extras add more source types:
+
+```bash
+pip install -e ".[pdf]"      # PDF (pypdf)
+pip install -e ".[docx]"     # Word documents
+pip install -e ".[media]"    # audio/video transcription (faster-whisper)
+pip install -e ".[youtube]"  # YouTube captions (yt-dlp)
+```
+
+## Methodology
+
+Five small stages, one traceable pipeline:
+
+1. **Ingest** — text/Markdown, web pages, PDF, DOCX, YouTube captions, audio and video (optional backends behind extras, imported lazily).
+2. **Chunk** — paragraph-aware splitting with overlap; media keeps start/end timestamps.
+3. **Note** — a provider drafts sections from chunks (offline extractive by default, or any OpenAI-compatible LLM with automatic fallback). Providers only return chunk ids; the core resolves them into citations and drops ids that do not exist, so hallucinated citations cannot leak into a note.
+4. **Ask** — dependency-free BM25-lite retrieval (CJK unigrams + bigrams, numeric tokens) finds the evidence; answers cite the chunks they came from.
+5. **Export** — Markdown for reading, Anki TSV for memorising.
+
+## Status and quality
+
+- 58 offline tests, 92% line coverage, CI on Python 3.9 / 3.11 / 3.12
+- Zero third-party dependencies in the core install
+- Bilingual: English and Chinese out of the box
+
+## Limitations
+
+- The default provider is extractive — it selects and organises sentences from your sources rather than rewriting them. Use an LLM provider for more fluent notes.
+- Retrieval is lexical (BM25): paraphrased questions can miss relevant passages; embeddings are on the roadmap.
+- Transcription quality depends on the ASR model and audio quality.
+- No live meeting capture by design: export the recording and upload it.
+- Notes can still miss context or nuance — always check important facts, numbers and quotes against the cited source.
+
+## More free Lynote tools
+
+- [📝 Free AI Note Taker](https://huggingface.co/spaces/Lynote/free-ai-notes) — hosted product page (lynote.ai/ai-notes)
+- [✍️ Free AI Humanizer](https://huggingface.co/spaces/Lynote/free-ai-humanizer) — rewrite AI text into natural human prose
+- [🔍 Free AI Detector](https://huggingface.co/spaces/Lynote/free-ai-detector) — sentence-level AI text detection
+- [🖼️ Free AI Image Detector](https://huggingface.co/spaces/Lynote/free-ai-image-detector) — check if an image was AI-generated
+- [🧠 humanize-text-model](https://huggingface.co/Lynote/humanize-text-model) — MIT-licensed bilingual T5 model
+
+## License
+
+MIT. See the [repository](https://github.com/lynote-ai/lynote-notes) for source and license notices.

app.py

@@ -0,0 +1,77 @@
+"""Interactive Lynote Notes demo (kept ready for PRO hosting).
+
+Paste text, get a cited note, then ask questions grounded in the same text.
+The demo uses the library's offline extractive provider, so it needs no model
+and no API key — only Gradio and the lynote-notes package.
+"""
+
+import tempfile
+
+import gradio as gr
+
+from lynote_notes import Workspace
+
+MAX_CHARS = 5_000
+UTM = (
+    "https://lynote.ai/ai-notes?utm_source=huggingface"
+    "&utm_medium=space&utm_campaign=hf_launch&utm_content=oss_notes"
+)
+
+
+def _workspace(text: str) -> Workspace:
+    ws = Workspace(tempfile.mkdtemp())
+    ws.add_text(text, "Pasted text")
+    return ws
+
+
+def make_note(text: str, title: str):
+    text = (text or "").strip()
+    if len(text) < 80:
+        return "Add at least 80 characters of source text."
+    if len(text) > MAX_CHARS:
+        return f"Input is limited to {MAX_CHARS:,} characters in this demo."
+    ws = _workspace(text)
+    note = ws.make_note(title or "Your note")
+    return ws.export_note(note, "md")
+
+
+def ask_question(text: str, question: str):
+    text = (text or "").strip()
+    question = (question or "").strip()
+    if not text or not question:
+        return "Paste some source text and ask a question."
+    ws = _workspace(text)
+    answer = ws.ask(question)
+    lines = [answer.text, ""]
+    for index, citation in enumerate(answer.citations, start=1):
+        lines.append(f"[{index}] {citation.source_title} · {citation.locator} — {citation.snippet}")
+    lines.append("")
+    lines.append(f"Try the full product: {UTM}")
+    return "\n".join(lines)
+
+
+with gr.Blocks(title="Lynote Notes — Source-Grounded AI Note Taker") as demo:
+    gr.Markdown(
+        "# 📝 Lynote Notes\n"
+        "Source-grounded notes with citations — every bullet links back to the exact "
+        "source chunk. Open source: [github.com/lynote-ai/lynote-notes]"
+        "(https://github.com/lynote-ai/lynote-notes)"
+    )
+    source = gr.Textbox(
+        label="Source text (80–5,000 characters)",
+        placeholder="Paste a lecture transcript, article, or meeting notes…",
+        lines=10,
+    )
+    with gr.Tab("Generate note"):
+        title = gr.Textbox(label="Note title", value="Your note")
+        note_button = gr.Button("Generate cited note", variant="primary")
+        note_output = gr.Markdown()
+        note_button.click(make_note, inputs=[source, title], outputs=note_output)
+    with gr.Tab("Ask a question"):
+        question = gr.Textbox(label="Question", placeholder="What did we decide about pricing?")
+        ask_button = gr.Button("Ask", variant="primary")
+        answer_output = gr.Markdown()
+        ask_button.click(ask_question, inputs=[source, question], outputs=answer_output)
+
+if __name__ == "__main__":
+    demo.launch()

index.html

@@ -1,19 +1,155 @@
-<!doctype html>
-<html>
-	<head>
-		<meta charset="utf-8" />
-		<meta name="viewport" content="width=device-width" />
-		<title>My static Space</title>
-		<link rel="stylesheet" href="style.css" />
-	</head>
-	<body>
-		<div class="card">
-			<h1>Welcome to your static Space!</h1>
-			<p>You can modify this app directly by editing <i>index.html</i> in the Files and versions tab.</p>
-			<p>
-				Also don't forget to check the
-				<a href="https://huggingface.co/docs/hub/spaces" target="_blank">Spaces documentation</a>.
-			</p>
-		</div>
-	</body>
+<!DOCTYPE html>
+<html lang="en">
+<head>
+<meta charset="UTF-8" />
+<meta name="viewport" content="width=device-width, initial-scale=1.0" />
+<title>Lynote Notes — Source-Grounded AI Note Taker (Open Source) | Lynote</title>
+<meta name="description" content="Open-source, source-grounded AI note taker: turn documents, web pages, audio, video and YouTube links into notes where every bullet cites the exact source chunk. Local-first, bilingual, zero-dependency core." />
+<meta name="keywords" content="open source ai note taker, source grounded notes, citations, rag notes, local first notes, ai notes cli, anki export, bilingual notes" />
+<meta property="og:title" content="Lynote Notes — Source-Grounded AI Note Taker (Open Source)" />
+<meta property="og:description" content="Notes where every bullet cites the exact source chunk. MIT licensed, local-first, bilingual." />
+<style>
+  :root { --bg:#0f1420; --card:#171f31; --card2:#1b2540; --fg:#e8edf7; --muted:#9aa7c0; --accent:#7b6dff; --accent2:#34d399; --line:#23304a; }
+  * { margin:0; padding:0; box-sizing:border-box; }
+  body { background:var(--bg); color:var(--fg); font-family:-apple-system,BlinkMacSystemFont,"Segoe UI","PingFang SC","Microsoft YaHei",sans-serif; line-height:1.7; }
+  .wrap { max-width:960px; margin:0 auto; padding:48px 24px 72px; }
+  .logo { font-size:15px; letter-spacing:3px; color:var(--muted); text-transform:uppercase; }
+  .badge { display:inline-block; font-size:13px; padding:4px 14px; border-radius:999px; border:1px solid #2c3a55; color:#b9c6e2; background:rgba(123,109,255,.08); margin:20px 0; }
+  h1 { font-size:clamp(30px,5.5vw,48px); line-height:1.2; margin:0 0 16px; background:linear-gradient(90deg,var(--accent),var(--accent2)); -webkit-background-clip:text; background-clip:text; -webkit-text-fill-color:transparent; }
+  .sub { font-size:18px; color:var(--muted); max-width:720px; }
+  .stats { display:grid; grid-template-columns:repeat(auto-fit,minmax(150px,1fr)); gap:14px; margin:34px 0 8px; }
+  .stat { background:var(--card); border:1px solid var(--line); border-radius:14px; padding:18px; text-align:center; }
+  .stat b { display:block; font-size:24px; background:linear-gradient(90deg,var(--accent),var(--accent2)); -webkit-background-clip:text; background-clip:text; -webkit-text-fill-color:transparent; }
+  .stat span { font-size:13px; color:var(--muted); }
+  h2 { font-size:26px; margin:56px 0 8px; }
+  .section-note { color:var(--muted); margin-bottom:22px; }
+  .grid { list-style:none; display:grid; grid-template-columns:repeat(auto-fit,minmax(260px,1fr)); gap:14px; }
+  .grid li { background:var(--card); border:1px solid var(--line); border-radius:14px; padding:18px 20px; }
+  .grid strong { display:block; color:#fff; margin-bottom:6px; font-size:16px; }
+  .grid span { color:var(--muted); font-size:14.5px; }
+  .steps { list-style:none; display:grid; grid-template-columns:repeat(auto-fit,minmax(220px,1fr)); gap:14px; margin-top:8px; }
+  .steps li { background:var(--card); border:1px solid var(--line); border-radius:14px; padding:20px; }
+  .steps em { display:block; font-style:normal; color:var(--accent2); font-size:13px; letter-spacing:1px; text-transform:uppercase; margin-bottom:6px; }
+  .steps strong { display:block; font-size:16px; margin-bottom:6px; }
+  .steps span { color:var(--muted); font-size:14.5px; }
+  pre { background:var(--card2); border:1px solid var(--line); border-radius:12px; padding:16px 18px; overflow-x:auto; font-size:13.5px; line-height:1.6; color:#d7e1f5; }
+  code { font-family:ui-monospace,SFMono-Regular,Menlo,Consolas,monospace; }
+  details { background:var(--card); border:1px solid var(--line); border-radius:12px; padding:14px 18px; margin-bottom:10px; }
+  summary { cursor:pointer; font-weight:600; color:#fff; }
+  details p { color:var(--muted); font-size:14.5px; margin-top:8px; }
+  .cta { display:inline-block; background:linear-gradient(90deg,var(--accent),var(--accent2)); color:#fff; font-size:17px; font-weight:600; padding:14px 30px; border-radius:12px; text-decoration:none; margin:36px 0 8px; }
+  .cta:hover { opacity:.9; }
+  .cta.secondary { background:var(--card); border:1px solid var(--line); font-weight:500; margin-left:10px; }
+  .links-label { font-size:12px; color:var(--muted); letter-spacing:1.5px; text-transform:uppercase; margin-top:36px; }
+  .links { display:flex; flex-wrap:wrap; gap:10px; align-items:center; margin-top:10px; }
+  .chip { font-size:14px; padding:8px 18px; border-radius:999px; border:1px solid var(--line); color:var(--fg); text-decoration:none; background:var(--card); }
+  .chip b { color:var(--accent2); }
+  .chip:hover { border-color:var(--accent); color:#fff; }
+  .note { font-size:13.5px; color:var(--muted); margin-top:24px; max-width:780px; }
+</style>
+</head>
+<body>
+<div class="wrap">
+
+  <div class="logo">Lynote</div>
+  <span class="badge">OPEN SOURCE · MIT · ZERO-DEPENDENCY CORE</span>
+  <h1>Lynote Notes</h1>
+  <p class="sub">
+    Source-grounded, bilingual, local-first AI note taker. Turn documents, web pages,
+    recordings, videos and YouTube links into structured notes where every bullet cites
+    the exact source chunk it came from — so you can verify, edit and reuse notes
+    instead of trusting a black box.
+  </p>
+
+  <a class="cta" href="https://github.com/lynote-ai/lynote-notes" target="_blank" rel="noopener">
+    View on GitHub →
+  </a>
+  <a class="cta secondary" href="https://lynote.ai/ai-notes?utm_source=huggingface&utm_medium=space&utm_campaign=hf_launch&utm_content=oss_notes" target="_blank" rel="noopener">
+    Try Lynote AI Notes
+  </a>
+
+  <div class="links-label">Free Lynote tools · open source</div>
+  <div class="links">
+    <a class="chip" href="https://github.com/lynote-ai/lynote-notes" target="_blank" rel="noopener"><b>GitHub</b> — lynote-ai/lynote-notes</a>
+    <a class="chip" href="https://huggingface.co/spaces/Lynote/free-ai-notes" target="_blank" rel="noopener">Free AI Note Taker →</a>
+    <a class="chip" href="https://huggingface.co/spaces/Lynote/free-ai-humanizer" target="_blank" rel="noopener">Free AI Humanizer →</a>
+    <a class="chip" href="https://huggingface.co/spaces/Lynote/free-ai-detector" target="_blank" rel="noopener">Free AI Detector →</a>
+    <a class="chip" href="https://huggingface.co/spaces/Lynote/free-ai-image-detector" target="_blank" rel="noopener">Free AI Image Detector →</a>
+  </div>
+
+  <div class="stats">
+    <div class="stat"><b>58</b><span>Tests Passing</span></div>
+    <div class="stat"><b>92%</b><span>Test Coverage</span></div>
+    <div class="stat"><b>0</b><span>Core Dependencies</span></div>
+    <div class="stat"><b>MIT</b><span>License</span></div>
+  </div>
+
+  <h2>Why Source-Grounded?</h2>
+  <p class="section-note">AI note tools usually give you a summary you cannot check. Lynote Notes is built around traceability.</p>
+  <ul class="grid">
+    <li><strong>Citations as a Contract</strong><span>Every bullet carries a source citation. Providers return chunk ids; the core resolves them and drops anything that does not exist — a hallucinated citation cannot leak into a note.</span></li>
+    <li><strong>Offline by Default</strong><span>The default extractive provider needs no model and no network. Configure any OpenAI-compatible endpoint (Ollama, vLLM, OpenAI) for generative notes, with automatic fallback.</span></li>
+    <li><strong>Bilingual by Design</strong><span>CJK unigrams + bigrams, sentence splitting on 。！？； — English and Chinese work out of the box.</span></li>
+    <li><strong>Local-First Workspace</strong><span>One workspace is one SQLite file. No server, no account, inspectable with standard tools.</span></li>
+    <li><strong>Media Timestamps</strong><span>Audio and video citations point to HH:MM:SS, not just "somewhere in the file".</span></li>
+    <li><strong>Reuse Built In</strong><span>Export Markdown for reading or Anki TSV for memorising — straight from the same cited note.</span></li>
+  </ul>
+
+  <h2>Quickstart</h2>
+  <p class="section-note">Python 3.9+, works offline.</p>
+  <pre><code>git clone https://github.com/lynote-ai/lynote-notes.git
+cd lynote-notes
+pip install -e .
+
+lynote-notes add lecture.md
+lynote-notes add https://example.com/post
+lynote-notes note --title "Weekly reading"
+lynote-notes ask "What did we decide about pricing?"
+lynote-notes export --note note_xxx --format anki --out cards.tsv</code></pre>
+
+  <h2>How It Works</h2>
+  <p class="section-note">Five small stages, one traceable pipeline.</p>
+  <ol class="steps">
+    <li><em>Step 1</em><strong>Ingest</strong><span>Text/Markdown, web pages, PDF, DOCX, YouTube captions, audio and video — optional backends stay behind extras.</span></li>
+    <li><em>Step 2</em><strong>Chunk</strong><span>Paragraph-aware splitting with overlap; media keeps timestamps.</span></li>
+    <li><em>Step 3</em><strong>Note</strong><span>A provider drafts sections from chunks; the core resolves every bullet to a real citation.</span></li>
+    <li><em>Step 4</em><strong>Ask</strong><span>BM25-lite retrieval finds the evidence; answers cite the chunks they came from.</span></li>
+    <li><em>Step 5</em><strong>Export</strong><span>Markdown for reading, Anki TSV for memorising.</span></li>
+  </ol>
+
+  <h2>Limitations</h2>
+  <ul class="grid">
+    <li><strong>Extractive Default</strong><span>The default provider selects and organises sentences from your sources rather than rewriting them. Use an LLM provider for more fluent notes.</span></li>
+    <li><strong>Lexical Retrieval</strong><span>BM25 matches words, not meanings. Paraphrased questions can miss relevant passages; embeddings are on the roadmap.</span></li>
+    <li><strong>Transcription Quality</strong><span>Depends on the ASR model size and audio quality — check important details against the source.</span></li>
+    <li><strong>No Live Capture</strong><span>No meeting bots by design: export the recording and upload it.</span></li>
+  </ul>
+
+  <h2>FAQ</h2>
+  <details>
+    <summary>Is Lynote Notes the same as Lynote AI Notes?</summary>
+    <p>Lynote Notes is the open-source core (MIT). Lynote AI Notes is the hosted product at lynote.ai/ai-notes, which adds a workspace UI, more formats and managed AI. The CLI and library here work standalone.</p>
+  </details>
+  <details>
+    <summary>Do I need an API key?</summary>
+    <p>No. The default extractive provider runs offline. An OpenAI-compatible endpoint is optional for generative notes.</p>
+  </details>
+  <details>
+    <summary>Does it work with Chinese content?</summary>
+    <p>Yes — tokenization and sentence splitting are bilingual by design, and notes can be generated from mixed-language workspaces.</p>
+  </details>
+  <details>
+    <summary>How do I verify a citation?</summary>
+    <p>Every citation names the source, a locator ("chunk 4" or "00:12:30") and a snippet. Open the source at that locator and compare — that is the point of the design.</p>
+  </details>
+
+  <p class="note">
+    This Space is a static demo page for the open-source project. The interactive Gradio demo
+    (paste text → cited note → ask questions) is kept ready in this repository and will be
+    enabled once the account has PRO hosting. Full product:
+    <a href="https://lynote.ai/ai-notes" style="color:var(--accent)">lynote.ai/ai-notes</a>.
+    AI notes can miss context or nuance — always check important details against the cited source.
+  </p>
+</div>
+</body>
 </html>

requirements.txt

@@ -0,0 +1,2 @@
+gradio>=4.0
+lynote-notes @ git+https://github.com/lynote-ai/lynote-notes.git@v0.1.0