<!--
  File:          web-landing/src/index.html
  Version:       1.9.0
  Last Modified: 2026-10-01 01:00 PM ET
  Changes:       1.9.0 (Task 41, One Collection): removed every "end-to-end", "zero-knowledge", "sealed on your
                 devices" and "Shared folder" sentence, the Vault Key / backup-code copy, and the shared-folder
                 asterisk fine print — one memory collection per user now, encrypted server-side with a per-user
                 key. Privacy section, features grid, pricing ($4.99) and FAQ rewritten to match
                 docs/one-collection-design.md §Copy.
                 1.8.1: Meta descriptions say private memories are end-to-end; asterisk on the kicker.
                 1.8.0: End-to-end encrypted for every account (privacy section, pricing, meta, FAQ). Cloud assistants use the
                 Shared folder: one asterisk + fine print. No "switch" language.
                 1.7.0: The live neuron brain is the first thing on the page, right under the menu; the headline follows it.
                 1.6.2: Wording: "zero-knowledge" replaced with "encrypted vault" until end-to-end encryption ships (the server still decrypts during requests).
                 1.6.1: Plain-language sales copy, complete feature coverage, clearer privacy limits, FAQ and metadata.
                 1.6.0: (rev) Accurate zero-knowledge wording: new accounts; the key is kept only sealed; server handles plaintext while saving/recalling.
                 Zero-knowledge by default section (with the honest fine print), full features grid (12), pricing lists
                 zero-knowledge, nav adds Zero-knowledge + Features. Old page backed up in backups/.
                 1.5.0: Hero shows the live neuron brain (10,000 example memories, demo assistants; brain-hero.js) instead of the static dashboard mockup; "Try the live demo" CTA.
                1.4.0: Final brand mark installed (Jeff's approved logo — brain + network-node mark, replacing
                 the provisional icon-a.svg concept that was pending approval). Added apple-touch-icon.
                 build.js no longer special-cases the favicon; it's copied from src/assets/ like any other asset.
                 1.3.0: copy fixes from docs/landing-copy-review.md — (1) the privacy bullets said memories are
                 "never" sent to a third party for embedding; the code also supports EMBEDDING_PROVIDER=openrouter,
                 so "never" wasn't true of the product, only of today's deployment config — reworded to say what's
                 actually true. (2) "34 assistant connections" -> "34 named assistant setups" (the registry has 35
                 client files; the 35th, "generic", isn't a named assistant, so 34 was already the right number for
                 what a visitor would count on the Connect page — just needed the word "named" to not read as a
                 total-coverage claim). (3) added the concrete differentiators the review found we have but weren't
                 saying: the live Verify check (Connect step), write-time duplicate detection (Remember step),
                 import support (Own step, next to the export it already mentioned), and an explicit no-annual-plan
                 line (pricing). Left the "Every assistant" hero headline and "Every assistant gets a tailored
                 setup" line as-is — softening those is a brand-voice call, not a correctness fix, and is Jeff's to
                 make deliberately rather than something to change quietly in a copy-accuracy pass.
                 1.2.0: font switched again, Inter -> Geist — Jeff wanted the CodePapi.net-adjacent look but
                 explicitly not Inter itself ("gets a lot of AI slop flack" — a real, current design-community
                 complaint that Inter signals "generic AI-generated SaaS template"). Geist (Vercel's open-source
                 grotesque, free, on Google Fonts) keeps the same clean/geometric family feel without that baggage.
                 1.1.0: headline font switched from Space Grotesk to Inter (matches codepapi.net's font choice,
                 per Jeff's request), loaded via Google Fonts so it actually renders everywhere (codepapi.net
                 itself never loads Inter — no @font-face/Google Fonts link and its CSP blocks external styles —
                 so it silently falls back to each visitor's OS font; this page loads the real thing).
                 Production landing page for MemoraMobi with CodePapi Yellow #FDC903 accent.
-->
<!doctype html>
<html lang="en">
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <meta name="description" content="One memory for all your AI assistants, encrypted with your own key. Free for 300 memories, Pro $4.99/month.">
  <title>MemoraMobi | Stop explaining yourself to every AI</title>
  <meta property="og:type" content="website">
  <meta property="og:url" content="https://memoramobi.com/">
  <meta property="og:title" content="Stop explaining yourself to every AI. | MemoraMobi">
  <meta property="og:description" content="One memory for all your AI assistants, encrypted with your own key. Free for 300 memories, Pro $4.99/month.">
  <meta property="og:site_name" content="MemoraMobi">
  <meta name="twitter:card" content="summary">
  <meta name="twitter:title" content="Stop explaining yourself to every AI. | MemoraMobi">
  <meta name="twitter:description" content="One memory for all your AI assistants, encrypted with your own key. Free for 300 memories, Pro $4.99/month.">
  
  <link rel="icon" type="image/svg+xml" href="/assets/favicon.svg">
  <link rel="apple-touch-icon" href="/assets/apple-touch-icon.png">
  
  <link rel="preconnect" href="https://fonts.googleapis.com">
  <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
  <link href="https://fonts.googleapis.com/css2?family=DM+Mono:wght@400;500&family=Geist:wght@400;600;700&display=swap" rel="stylesheet">
  <link rel="stylesheet" href="/assets/styles.css">
  <link rel="stylesheet" href="/assets/brain-hero.css">
  <script defer src="/assets/brain-hero.js"></script>
</head>
<body>
  <header class="wrap nav">
    <a class="brand" href="#top">
      <img src="/assets/favicon.svg" alt="MemoraMobi" class="brand-icon">
      Memora<b>Mobi</b>
    </a>
    <nav aria-label="Main navigation">
      <a href="#how">How it works</a>
      <a href="#privacy">Privacy</a>
      <a href="#features">Features</a>
      <a href="#pricing">Pricing</a>
      <a href="/docs">MCP Docs</a>
    </nav>
    <a class="cta" href="https://app.memoramobi.com">Start free</a>
  </header>

  <main id="top">
    <section class="wrap hero">
      <div class="brain-hero" id="mm-brain-hero" role="img" aria-label="Live demo: a MemoraMobi memory brain with 10,000 example memories and demo assistants saving and recalling">
        <div class="brain-hero-fallback">
          <span class="mono">LIVE DEMO</span>
          <b>See a memory travel between assistants</b>
          <p>This demo uses example memories. In your account, Mission Control shows what your connected assistants save and recall as it happens.</p>
        </div>
      </div>
      <div class="brain-hero-cta">
        <a class="button primary" href="https://app.memoramobi.com/try">Try the live demo →</a>
        <span class="mono">10,000 EXAMPLE MEMORIES · RESETS HOURLY · LOGIN: demo / demo</span>
      </div>

      <div class="kicker">
        <span class="mono">ONE MEMORY FOR YOUR AI ASSISTANTS</span>
        <span class="signal"><i></i> 300 MEMORIES FREE</span>
      </div>

      <h1>Stop explaining yourself<br><em>to every AI.</em></h1>

      <div class="hero-lede">
        <p>Stop repeating your preferences and project details. MemoraMobi gives Claude, ChatGPT, Cursor, Codex, Gemini and more a shared memory, so the next assistant can pick up where you left off.</p>
        <div class="buttons">
          <a class="button primary" href="https://app.memoramobi.com">Start free →</a>
          <a class="button" href="https://app.memoramobi.com/try">Try it without signing up</a>
        </div>
      </div>

    </section>

    <section class="section" id="how">
      <div class="wrap">
        <span class="mono">HOW IT WORKS</span>
        <h2>Tell one assistant.<br>Let the next one remember.</h2>
        <p class="section-intro">Tell Claude you’re vegetarian and ask it to remember. Later, ask your connected ChatGPT for dinner ideas. It can recall that preference from MemoraMobi, so you can get straight to dinner.</p>
        <div class="timeline">
          <div class="step">
            <b>01</b>
            <h3>Pick your assistant</h3>
            <p>Choose from 34 named assistants. Follow its setup guide, use one-click install where supported, then check that the connection worked.</p>
          </div>
          <div class="step">
            <b>02</b>
            <h3>Save what matters</h3>
            <p>Your AI chooses which facts to save. MemoraMobi stores those facts, not raw chat transcripts, and checks for exact and near-duplicates.</p>
          </div>
          <div class="step">
            <b>03</b>
            <h3>Change assistants</h3>
            <p>The next connected assistant can look up the details it needs. Meaning-based and keyword search help it find a memory even if you phrase things differently.</p>
          </div>
          <div class="step">
            <b>04</b>
            <h3>Keep it yours</h3>
            <p>Read and edit your saved memories whenever you like. Bring old memories with you, export everything, or forget a detail with undo if you change your mind.</p>
          </div>
        </div>
      </div>
    </section>

    <section class="section" id="privacy">
      <div class="wrap privacy">
        <div class="privacy-copy">
          <span class="mono">ENCRYPTED WITH YOUR OWN KEY</span>
          <h2>Encrypted with your own key.</h2>
          <p>Every memory — text, tags and search data — is stored encrypted with a key that is yours alone. A stolen copy of our database or backups can't read it. The key is kept by a locked-down service on our server and used only while you or an assistant you approved is using your memories. Assistants receive the memories they ask for; disconnect one and its access ends.</p>
          <p class="footnote">MemoraMobi's service can technically open your memories during your requests — that is what lets ChatGPT, Claude and the others use them without installing anything. We don't read them, we log only ids and counts, and we say so here instead of using words like zero-knowledge that wouldn't be true.</p>
        </div>
        <div class="privacy-card">
          <strong>You decide who has access.</strong>
          <ul>
            <li>See your connected apps and disconnect any one in a click. Its access ends there.</li>
            <li>Search runs inside the same locked-down service, on your encrypted data. We never send memory text to OpenAI, Google or another AI company for that work.</li>
            <li>Your connected assistant receives the memories it recalls. It already handles the conversation you have with it.</li>
            <li>We store only the facts your AI chooses to save. No raw chat transcripts.</li>
            <li>You can export your memories or delete your account and everything in it.</li>
          </ul>
        </div>
      </div>
    </section>

    <section class="section" id="features">
      <div class="wrap">
        <span class="mono">BUILT FOR EVERYDAY USE</span>
        <h2>Less repeating.<br>More getting things done.</h2>
        <div class="features">
          <div class="feature"><h3>Keep the assistants you like</h3><p>34 named assistants, including Claude, ChatGPT, Cursor, Codex, Gemini, VS Code and Windsurf. Each has a tailored setup guide and a live connection check.</p></div>
          <div class="feature"><h3>Connect without copying a key</h3><p>Claude.ai and ChatGPT connectors support “Sign in with MemoraMobi” through OAuth. Approve once. Manage access in Connected apps, with one-click Disconnect.</p></div>
          <div class="feature"><h3>Watch your memories at work</h3><p>Mission Control is a live 3D neuron brain. Saves, recalls and merges light up as they happen. Follow the activity feed and charts, or click a group to open it.</p></div>
          <div class="feature"><h3>See which assistant is using it</h3><p>The live Connections map shows saves flowing in and recalls flowing out for each assistant. A handshake animation marks a new connection.</p></div>
          <div class="feature"><h3>Find things where you expect</h3><p>Memories sort into people, work, health… on our servers, encrypted at rest.</p></div>
          <div class="feature"><h3>Keep up when life changes</h3><p>New address? New job? Newer facts replace older ones through a merge. Every edit keeps its version history, so you can restore an earlier version.</p></div>
          <div class="feature"><h3>Catch repeats early</h3><p>Exact and near-duplicate detection helps keep the same fact from piling up each time an assistant saves it.</p></div>
          <div class="feature"><h3>Find the detail you meant</h3><p>Smart recall combines meaning-based and keyword search. You don’t have to remember the exact words you used to save something.</p></div>
          <div class="feature"><h3>Bring what your AI already knows</h3><p>Guided import brings memories over from ChatGPT, Claude and other AIs, with duplicate detection. Export everything any time.</p></div>
          <div class="feature"><h3>Use it in your own tools</h3><p>The Model Context Protocol (MCP) connects your assistants. A plain REST API uses the same keys for scripts, custom GPT Actions, automations and your own apps.</p></div>
          <div class="feature"><h3>Same memory, every assistant</h3><p>Works in every assistant: ChatGPT, Claude.ai, Claude on your phone, Claude Code, Cursor, Codex, Gemini CLI — same memories everywhere. Tools without a connector get a “memory brief” you paste in.</p></div>
          <div class="feature"><h3>Change your mind</h3><p>Edit a memory, regroup it, or forget it with undo. You can also delete your account and everything in it. You stay in charge of what’s saved.</p></div>
        </div>
      </div>
    </section>

    <section class="section" id="pricing">
      <div class="wrap">
        <span class="mono">PRICING</span>
        <h2>Start free.<br>Pay when you need more room.</h2>
        <div class="price">
          <article class="price-card">
            <h3>Free</h3>
            <div class="amount">$0</div>
            <p>300 memories. Every feature included.</p>
            <ul>
              <li>300 memories</li>
              <li>All 34 assistant setup guides</li>
              <li>Encrypted with your own key</li>
              <li>Mission Control, Connections, history, import and export</li>
            </ul>
            <a class="button" href="https://app.memoramobi.com">Start free</a>
          </article>
          <article class="price-card featured">
            <h3>Pro</h3>
            <div class="amount">$4.99<small>/mo</small></div>
            <p>Keep saving without counting memories.</p>
            <ul>
              <li>Unlimited memories</li>
              <li>Everything in Free, encrypted the same way</li>
              <li>Monthly billing. Cancel any time.</li>
            </ul>
            <a class="button primary" href="https://app.memoramobi.com">Choose Pro</a>
          </article>
        </div>
        <p class="footnote">Free includes every feature, encrypted with your own key. Pro is $4.99 per month for unlimited memories. No annual plan. Cancel any time.</p>
      </div>
    </section>
    <section class="section" id="questions">
      <div class="wrap">
        <span class="mono">A FEW THINGS YOU MIGHT BE WONDERING</span>
        <h2>Before you connect.</h2>
        <div class="faq">
          <details><summary>Will it remember every conversation automatically?</summary><p>No. Your assistant chooses which facts to save and when to recall them. MemoraMobi stores those facts, not a recording of your chats. The setup guide helps you give your assistant instructions for using memory.</p></details>
          <details><summary>Can I try it before making an account?</summary><p>Yes. <a href="https://app.memoramobi.com/try">Open the public demo</a> with username <code>demo</code> and password <code>demo</code>. Explore 10,000 example memories. It resets hourly, so use examples rather than personal details.</p></details>
          <details><summary>Do I need to run a server?</summary><p>No. MemoraMobi is hosted on our servers. Tools without a remote MCP connector use a small local adapter (one line in their settings) that talks to our servers. A self-hosted option is coming later.</p></details>
          <details><summary>Is this zero-knowledge?</summary><p>No. Zero-knowledge would mean we could never open your memories, and then ChatGPT or Claude couldn't use them either. We store everything encrypted and open it only during your requests.</p></details>
          <details><summary>What if I lose my passphrase?</summary><p>Reset it by email, same as any account. Nothing is lost.</p></details>
          <details><summary>What if I stop using MemoraMobi?</summary><p>Export everything any time. You can cancel Pro whenever you like, or delete your account and everything in it.</p></details>
        </div>
        <div class="callout">
          <div><h3>Your next assistant doesn’t have to start from scratch.</h3><p>Start with 300 free memories. Connect an assistant and save your first fact.</p></div>
          <a class="button primary" href="https://app.memoramobi.com">Start free →</a>
        </div>
      </div>
    </section>
  </main>

  <footer class="wrap">
    <span>© 2026 MemoraMobi · Take your memories to your next assistant.</span>
    <span class="footer-links">
      Hosted on MemoraMobi servers · <a href="/docs">MCP Docs</a> · <a href="https://app.memoramobi.com/connect">Connect an assistant</a>
    </span>
  </footer>
</body>
</html>

<!-- === END OF FILE (FILENAME: /opt/memoria-mobilis/web-landing/src/index.html) === -->