fix(launch): reliable Windows launcher; feat(ui): compact sidebar, Vite toggle, chat Think toggle

Ported from the private repo via bin/publish.sh, plus a manual catch-up on files that had drifted out of sync before today: - launch_nexus.ps1: health-check based restart decisions instead of a bare port-listen check (a wedged leftover process squatting a port used to look "already running" and block the real service from starting), a script-path quoting fix for Start-Process, hidden console via a wscript.exe wrapper (bin/launch_nexus_hidden.vbs), and a taskbar/window icon for the native app window. - Sidebar: slim icon+text nav rows instead of bulky bordered buttons, tighter spacing throughout. - Settings: full-width layout, a Vite dev-server Start/Stop toggle (synapse/frontend_manager.py + /frontend/* endpoints), and the Linux-only Icon Branding section now gated on the new /status `platform` field instead of always rendering. - Chatbot: a Think toggle next to the model picker, so extended thinking can be flipped without leaving the chat page. - management/ncp.py: faster start/stop polling (0.25s steps instead of 1s), Vite no longer blocks `ncp start` on Linux and is skipped outright on Windows. Note: the private repo also has a Mail (IMAP/SMTP) feature; it's intentionally not included here, so the Mail-only pieces of main.py, App.jsx, and requirements-windows.txt were left out of this port.
2026-07-28 11:23:35 -05:00
parent 63c93346ae
commit cc20ceac64
11 changed files with 490 additions and 109 deletions
@@ -153,6 +153,43 @@ Promethean/       Python venv (gitignored, built by the installer)
 | Frontend API base URL | `interface/web/src/config.js` |
 | Python deps | `requirements-base.txt` + amd/nvidia GPU overlay; `requirements-windows.txt` = standalone CPU runtime |

+## Roadmap
+
+Rough plan, not promises. Ordered by how soon and how much it touches.
+
+### Soon
+- **Inference knobs in Settings** — expose temperature, context length, and
+  `num_gpu` in the UI. The backend already passes options to Ollama; there's no
+  UI for them yet, and small-VRAM boxes need `num_gpu=0` today via env only.
+- **Chat controls** — stop generation, regenerate last reply, edit-and-resend.
+  History editing exists; live regeneration doesn't.
+- **Per-playbook model** — playbooks set the persona but not the model; let a
+  playbook pin its own chat model (Settings already splits chat vs. memory model).
+
+### Eventually
+- **Document ingest / RAG** — the embedding stack (nomic-embed-text,
+  `message_vectors`, cosine recall) already retrieves past conversations. Extend
+  it to uploaded files and notes so chats can cite your own documents.
+- **Vision chat** — attach images to a message and route to a multimodal Ollama
+  model. The stream path is text-only today.
+- **Real vector index** — recall does a brute-force cosine scan over JSON blobs
+  in SQLite. Fine at current scale; swap in a proper index (e.g. sqlite-vec)
+  before the DB grows.
+
+### Far future
+- **macOS support** — POSIX process control already works; needs a `sync.py`
+  platform fork, ROCm-free requirements, and an installer. Waiting on a Mac.
+- **Voice I/O** — local speech-to-text in, text-to-speech out.
+- **Fine-tuning loop** — conversations already export as ShareGPT; close the
+  loop into a local fine-tune. The ML stack was stripped from the runtime venv
+  but the requirements files restore it.
+- **Tool-using playbooks** — function calling so a playbook can take actions,
+  not just shape the prompt.
+- **Machine learning & gameplay** — the original goal: an agent that learns to
+  play games alongside me (it started as "teach it to play Lego Star Wars").
+  Screen capture in, controller/input out, trained by playing. The whole reason
+  this project exists.
+
 ---

 <div align="center"><sub>NexusOS · local AI, self-hosted on <a href="https://git.enderofwings.com/enderofwings/NexusOS">GitNexus</a></sub></div>
@@ -0,0 +1,15 @@
+' Truly-hidden entry point for the NexusOS desktop shortcut.
+'
+' powershell.exe -WindowStyle Hidden is not enough on Windows 11: when Windows
+' Terminal is the registered default terminal app, it intercepts console-
+' subsystem processes before the hide flag applies, so a "NexusOS" terminal
+' tab shows up anyway. wscript.exe is a GUI-subsystem host (unlike its sibling
+' cscript.exe), so it never allocates a console in the first place and is not
+' a candidate for that hijack - WScript.Shell.Run's own hidden window style
+' (0) is then respected for the powershell.exe it launches.
+Dim fso, scriptDir, psPath, cmd
+Set fso = CreateObject("Scripting.FileSystemObject")
+scriptDir = fso.GetParentFolderName(WScript.ScriptFullName)
+psPath = fso.GetParentFolderName(scriptDir) & "\launch_nexus.ps1"
+cmd = "powershell.exe -NoProfile -ExecutionPolicy Bypass -WindowStyle Hidden -File """ & psPath & """"
+CreateObject("WScript.Shell").Run cmd, 0, False
@@ -10,19 +10,27 @@ The window opens IMMEDIATELY on a loading page and navigates to the app once the
 backend answers, rather than waiting up to 40s with nothing on screen and then
 appearing. One window the whole time: nothing pops up and vanishes, and there is
 no moment where the user is left wondering whether the command did anything.
+The loading page's status line is updated live (memory -> backend -> ready) via
+evaluate_js, instead of sitting on one static sentence for the whole wait - the
+point is a user watching the window can tell what stage it's stuck on, without
+needing to go read a terminal or a log file to find out.

 Falls back to the default browser if pywebview/WebView2 is unavailable, staying
 alive so the launcher doesn't tear the services down underneath it.
 """
+import os
 import sys
 import time
 import urllib.request

 URL = "http://localhost:8000"
+MEMORY_URL = "http://localhost:8001/"
+ICON_PATH = os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), "assets", "NexusOS.ico")

 # Inline so it renders with no server and no asset files - the whole point is
 # that it shows before anything is listening. Colours match the app's dark UI so
-# the swap to the real page is not a flash of a different-looking window.
+# the swap to the real page is not a flash of a different-looking window. The
+# <p id> is the hook _set_status() rewrites as boot moves through its stages.
 LOADING_HTML = """<!doctype html>
 <html><head><meta charset="utf-8"><title>NexusOS</title><style>
  html,body{height:100%;margin:0}
@@ -38,27 +46,57 @@ LOADING_HTML = """<!doctype html>
 </style></head><body>
  <div class="box">
    <div class="ring"></div>
-    <h1>Loading Nexus core services</h1>
-    <p>Starting the memory service and backend...</p>
+    <h1>Starting NexusOS</h1>
+    <p id="status">Starting memory service...</p>
  </div>
 </body></html>"""

 FAILED_HTML = LOADING_HTML.replace(
    "<div class=\"ring\"></div>", ""
 ).replace(
-    "Loading Nexus core services", "Backend did not start"
+    "Starting NexusOS", "Backend did not start"
 ).replace(
-    "Starting the memory service and backend...",
-    "Nothing answered on :8000 after 40s. Check: ncp logs -b"
+    'id="status">Starting memory service...',
+    'id="status">Nothing answered on :8000 after 40s. Check: ncp logs -b',
 )


-def _wait_for_backend(timeout: float = 40.0) -> bool:
-    """Poll /status until the backend answers, so the window never loads before
-    the server is up (which shows a localhost error the webview won't retry)."""
+def _set_status(win, text: str) -> None:
+    # Best-effort: the window can be mid-teardown (user closed it while this
+    # background thread was still polling), and evaluate_js on a dead window
+    # raises - that's not a reason to kill the polling loop.
+    try:
+        win.evaluate_js(f"document.getElementById('status').textContent = {text!r}")
+    except Exception:
+        pass
+
+
+def _http_ok(url: str, timeout: float = 2.0) -> bool:
+    try:
+        with urllib.request.urlopen(url, timeout=timeout) as r:
+            return r.status == 200
+    except Exception:
+        return False
+
+
+def _wait_for_backend(win, timeout: float = 40.0) -> bool:
+    """Poll memory, then the backend, updating the on-screen status as each
+    comes up - so the window never loads before the server is up (which shows
+    a localhost error the webview won't retry), and never sits on one sentence
+    while actually moving through two separate services."""
    status_url = URL.rstrip("/") + "/status"
    deadline = time.time() + timeout
+
+    memory_ready = _http_ok(MEMORY_URL)
+    if memory_ready:
+        _set_status(win, "Starting backend...")
+
    while time.time() < deadline:
+        if not memory_ready:
+            memory_ready = _http_ok(MEMORY_URL)
+            if memory_ready:
+                _set_status(win, "Starting backend...")
+
        try:
            # 5s, not 2: /status probes Ollama, and a wedged Ollama made it
            # slower than a 2s ceiling - the backend was up and answering 200
@@ -66,9 +104,11 @@ def _wait_for_backend(timeout: float = 40.0) -> bool:
            # The server side is fixed too; this is the margin.
            with urllib.request.urlopen(status_url, timeout=5) as r:
                if r.status == 200:
+                    _set_status(win, "Opening Nexus...")
                    return True
        except Exception:
-            time.sleep(1)
+            pass
+        time.sleep(0.5)
    return False


@@ -91,14 +131,15 @@ def main() -> int:
            """Runs once the GUI loop is up, so the spinner is already on screen
            while we wait. load_url replaces the loading page in place - there is
            never a second window to close."""
-            if _wait_for_backend():
+            if _wait_for_backend(win):
                win.load_url(URL)
            else:
                print("[nexus] backend not reachable on :8000 after 40s.",
                      file=sys.stderr)
                win.load_html(FAILED_HTML)

-        webview.start(_swap_in_app, window)  # blocks until the window is closed
+        icon = ICON_PATH if os.path.isfile(ICON_PATH) else None
+        webview.start(_swap_in_app, window, icon=icon)  # blocks until the window is closed
        return 0
    except Exception as e:  # no WebView2 runtime / backend failure
        return _browser_fallback(f"native window failed ({e})")
@@ -217,12 +217,16 @@ if ($Shadowed) {
 # -- Desktop shortcut ----------------------------------------------------------
 Write-Step "Creating desktop shortcut"
 try {
-    $Launcher = Join-Path $RepoRoot "launch_nexus.ps1"
+    # Target the .vbs wrapper, not powershell.exe directly: on Windows 11 with
+    # Windows Terminal as the default terminal app, -WindowStyle Hidden on a
+    # console-subsystem process is often ignored and a terminal tab flashes up
+    # anyway. wscript.exe never allocates a console at all, sidestepping that.
+    $Launcher = Join-Path $RepoRoot "bin\launch_nexus_hidden.vbs"
    $LnkPath = Join-Path ([Environment]::GetFolderPath("Desktop")) "NexusOS.lnk"
    $ws = New-Object -ComObject WScript.Shell
    $lnk = $ws.CreateShortcut($LnkPath)
-    $lnk.TargetPath = "powershell.exe"
-    $lnk.Arguments = "-NoProfile -ExecutionPolicy Bypass -WindowStyle Hidden -File `"$Launcher`""
+    $lnk.TargetPath = "$env:WINDIR\System32\wscript.exe"
+    $lnk.Arguments = "`"$Launcher`""
    $lnk.WorkingDirectory = $RepoRoot
    $lnk.Description = "Launch NexusOS"
    $Ico = Join-Path $RepoRoot "assets\NexusOS.ico"
@@ -155,20 +155,20 @@ function App() {
  };

  const navItems = [
-    { key: "chatbot", label: "💬 Chat" },
-    { key: "playbook", label: "📖 Playbooks" },
-    { key: "models", label: "🤖 Models", badge: isModelPulling },
-    { key: "memory", label: "🧠 Memory" },
-    { key: "documents", label: "📄 Documents" },
-    { key: "logs", label: "📜 Logs" },
-    { key: "settings", label: "⚙️ Settings" },
+    { key: "chatbot", icon: "💬", label: "Chat" },
+    { key: "playbook", icon: "📖", label: "Playbooks" },
+    { key: "models", icon: "🤖", label: "Models", badge: isModelPulling },
+    { key: "memory", icon: "🧠", label: "Memory" },
+    { key: "documents", icon: "📄", label: "Documents" },
+    { key: "logs", icon: "📜", label: "Logs" },
+    { key: "settings", icon: "⚙️", label: "Settings" },
  ];

  return (
    <div style={{ background: "#111", color: "#eee", height: "100vh", overflow: "hidden", fontFamily: "system-ui", display: "flex" }}>
      {/* Sidebar */}
      <aside style={{
-        width: "260px",
+        width: "220px",
        flexShrink: 0,
        background: "#0a0a0a",
        borderRight: "1px solid #333",
@@ -178,18 +178,18 @@ function App() {
      }}>
        {/* Top cell: brand + nav */}
        <div style={{
-          padding: "1.25rem 1rem 1rem",
+          padding: "0.75rem 0.6rem 0.65rem",
          borderBottom: "1px solid #1f1f1f",
          display: "flex",
          flexDirection: "column",
          alignItems: "center",
-          gap: "0.75rem",
+          gap: "0.4rem",
          flexShrink: 0,
        }}>
-          <img src="/n small.png" alt="Logo" style={{ width: "48px", height: "48px", objectFit: "contain", borderRadius: "8px" }} />
-          <h1 style={{ fontSize: "1rem", margin: 0, color: "#007acc", textAlign: "center" }}>NexusOS</h1>
+          <img src="/n small.png" alt="Logo" style={{ width: "36px", height: "36px", objectFit: "contain", borderRadius: "6px" }} />
+          <h1 style={{ fontSize: "0.9rem", margin: 0, color: "#007acc", textAlign: "center" }}>NexusOS</h1>
          {version && (
-            <span style={{ fontSize: "0.7rem", color: "#666", marginTop: "-0.5rem", letterSpacing: "0.02em" }}>v{version}</span>
+            <span style={{ fontSize: "0.65rem", color: "#666", marginTop: "-0.35rem", letterSpacing: "0.02em" }}>v{version}</span>
          )}

          {/* Status Indicator */}
@@ -243,13 +243,13 @@ function App() {
            title={ollamaStatus === "unavailable" ? "Ollama binary not found" : "Start or stop the AI"}
            style={{
              width: "100%",
-              padding: "0.5rem 0.75rem",
+              padding: "0.4rem 0.6rem",
              background: ollamaStatus === "running" ? "#2a1a1a" : "#152a15",
              color: ollamaStatus === "running" ? "#ff8a80" : "#8aff8a",
              border: "1px solid " + (ollamaStatus === "running" ? "#5a2a2a" : "#2a5a2a"),
-              borderRadius: "8px",
+              borderRadius: "6px",
              cursor: (ollamaBusy || ollamaStatus === "unavailable" || status === "Offline") ? "not-allowed" : "pointer",
-              fontSize: "0.8rem",
+              fontSize: "0.78rem",
              opacity: (ollamaStatus === "unavailable" || status === "Offline") ? 0.5 : 1,
            }}
          >
@@ -260,39 +260,46 @@ function App() {
              : ollamaStatus === "running" ? "⏹ Stop AI" : "▶ Start AI"}
          </button>

-          <nav style={{ display: "flex", flexDirection: "column", gap: "0.4rem", width: "100%", marginTop: "0.25rem" }}>
-            {navItems.map(item => (
-              <button
-                key={item.key}
-                onClick={() => setCurrentPage(item.key)}
-                style={{
-                  padding: "0.65rem 0.85rem",
-                  background: currentPage === item.key ? "#007acc" : "#161616",
-                  color: "#fff",
-                  border: "1px solid " + (currentPage === item.key ? "#0099ff" : "#2a2a2a"),
-                  borderRadius: "8px",
-                  cursor: "pointer",
-                  fontSize: "0.85rem",
-                  textAlign: "left",
-                  transition: "all 0.15s",
-                  display: "flex",
-                  alignItems: "center",
-                  justifyContent: "space-between",
-                }}
-              >
-                {item.label}
-                {item.badge && (
-                  <span style={{
-                    width: "8px",
-                    height: "8px",
-                    borderRadius: "50%",
-                    background: "#28a745",
-                    flexShrink: 0,
-                    animation: "pulse 1.2s ease-in-out infinite",
-                  }} />
-                )}
-              </button>
-            ))}
+          <nav style={{ display: "flex", flexDirection: "column", width: "100%", marginTop: "0.1rem" }}>
+            {navItems.map(item => {
+              const isActive = currentPage === item.key;
+              return (
+                <button
+                  key={item.key}
+                  onClick={() => setCurrentPage(item.key)}
+                  style={{
+                    padding: "0.32rem 0.5rem",
+                    background: isActive ? "#1c2733" : "transparent",
+                    color: isActive ? "#fff" : "#bbb",
+                    border: "none",
+                    borderRadius: "5px",
+                    cursor: "pointer",
+                    fontSize: "0.8rem",
+                    textAlign: "left",
+                    transition: "background 0.12s, color 0.12s",
+                    display: "flex",
+                    alignItems: "center",
+                    gap: "0.45rem",
+                    width: "100%",
+                  }}
+                  onMouseEnter={(e) => { if (!isActive) e.currentTarget.style.background = "#161616"; }}
+                  onMouseLeave={(e) => { if (!isActive) e.currentTarget.style.background = "transparent"; }}
+                >
+                  <span style={{ fontSize: "0.85rem", width: "1rem", textAlign: "center", flexShrink: 0 }}>{item.icon}</span>
+                  <span style={{ flexGrow: 1 }}>{item.label}</span>
+                  {item.badge && (
+                    <span style={{
+                      width: "6px",
+                      height: "6px",
+                      borderRadius: "50%",
+                      background: "#28a745",
+                      flexShrink: 0,
+                      animation: "pulse 1.2s ease-in-out infinite",
+                    }} />
+                  )}
+                </button>
+              );
+            })}
          </nav>
        </div>

@@ -302,7 +309,7 @@ function App() {
          minHeight: 0,
          display: "flex",
          flexDirection: "column",
-          padding: "0.75rem 0.75rem 0.75rem",
+          padding: "0.6rem 0.6rem 0.6rem",
        }}>
          <div style={{ display: "flex", justifyContent: "space-between", alignItems: "center", marginBottom: "0.5rem", flexShrink: 0 }}>
            <span style={{ fontSize: "0.75rem", color: "#888", textTransform: "uppercase", letterSpacing: "0.05em" }}>Chats</span>
@@ -470,13 +477,19 @@ function App() {
        }}>
          <Models onPullStateChange={setIsModelPulling} />
        </div>
-        {currentPage === "chatbot" && (
+        {/* Chatbot stays mounted so an in-flight reply survives page navigation */}
+        <div style={{
+          display: currentPage === "chatbot" ? "flex" : "none",
+          flexDirection: "column",
+          flexGrow: 1,
+          minHeight: 0,
+        }}>
          <Chatbot
            conversationId={activeConversationId}
            setConversationId={setActiveConversationId}
            onConversationChanged={() => loadConversations(search)}
          />
-        )}
+        </div>
        {currentPage === "playbook" && <Playbook />}
        {currentPage === "memory" && <Memory />}
        {currentPage === "documents" && <Documents />}
@@ -10,6 +10,7 @@ export function Chatbot({ conversationId, setConversationId, onConversationChang
  const [modelList, setModelList] = useState([]);
  const [selectedModel, setSelectedModel] = useState(""); // "" = auto
  const [autoModel, setAutoModel] = useState(null);
+  const [think, setThink] = useState(false); // extended thinking, mirrors Settings
  const [showPicker, setShowPicker] = useState(false);
  const [copiedIdx, setCopiedIdx] = useState(null);
  const [lastStats, setLastStats] = useState(null);
@@ -138,7 +139,10 @@ export function Chatbot({ conversationId, setConversationId, onConversationChang
    ]).then(([models, settings]) => {
      if (models?.models) setModelList(models.models);
      if (models?.selected) setAutoModel(models.selected);
-      if (settings) setSelectedModel(settings.model || "");
+      if (settings) {
+        setSelectedModel(settings.model || "");
+        setThink(!!settings.think);
+      }
    }).catch(() => {});
  }, []);

@@ -163,6 +167,18 @@ export function Chatbot({ conversationId, setConversationId, onConversationChang
    } catch { /* persisting the model choice is best-effort */ }
  };

+  const toggleThink = async () => {
+    const next = !think;
+    setThink(next);
+    try {
+      await fetch(`${API_BASE}/settings`, {
+        method: "PUT",
+        headers: { "Content-Type": "application/json" },
+        body: JSON.stringify({ think: next }),
+      });
+    } catch { /* persisting the toggle is best-effort */ }
+  };
+
  const startNewChat = () => {
    if (abortRef.current) abortRef.current.abort();
    setInput("");
@@ -202,6 +218,7 @@ export function Chatbot({ conversationId, setConversationId, onConversationChang
          message,
          conversation_id: conversationId,
          history,
+          think,
          ...(imgs && imgs.length ? { images: imgs } : {}),
        }),
        signal: controller.signal,
@@ -522,6 +539,28 @@ export function Chatbot({ conversationId, setConversationId, onConversationChang
                </div>
              )}
            </div>
+            <button
+              onClick={toggleThink}
+              title={think ? "Extended thinking is on — click to turn off" : "Extended thinking is off — click to turn on"}
+              style={{
+                fontSize: "0.7rem",
+                color: think ? "#c4b5fd" : "#888",
+                background: think ? "#1e1a2e" : "#1a1a1a",
+                border: "1px solid " + (think ? "#7c3aed" : "#2a2a2a"),
+                borderRadius: "4px",
+                padding: "0.15rem 0.5rem",
+                cursor: "pointer",
+                letterSpacing: "0.03em",
+                display: "flex",
+                alignItems: "center",
+                gap: "0.3rem",
+              }}
+            >
+              🧠 Think
+              <span style={{ fontSize: "0.6rem", color: think ? "#8aff8a" : "#555" }}>
+                {think ? "on" : "off"}
+              </span>
+            </button>
            {lastStats && (
              <span style={{
                fontSize: "0.7rem",
@@ -33,6 +33,34 @@ export function Settings() {
  const [brandQueue, setBrandQueue] = useState([]);
  const [isBranding, setIsBranding] = useState(false);
  const [modelList, setModelList] = useState([]);
+  const [platform, setPlatform] = useState("");
+
+  // Vite dev server toggle
+  const [viteRunning, setViteRunning] = useState(null);
+  const [viteBusy, setViteBusy] = useState(false);
+  const [viteError, setViteError] = useState("");
+
+  const loadViteStatus = () => {
+    fetch(`${API_BASE}/frontend/status`)
+      .then(r => r.ok ? r.json() : null)
+      .then(d => { if (d) setViteRunning(!!d.running); })
+      .catch(() => {});
+  };
+
+  const toggleVite = async () => {
+    setViteBusy(true);
+    setViteError("");
+    try {
+      const action = viteRunning ? "stop" : "start";
+      const r = await fetch(`${API_BASE}/frontend/${action}`, { method: "POST" });
+      const d = await r.json().catch(() => null);
+      if (d && d.status === "error") setViteError(d.detail || "Failed to start Vite");
+    } catch (e) {
+      setViteError(e.message || "Failed");
+    }
+    setViteBusy(false);
+    loadViteStatus();
+  };

  useEffect(() => {
    fetch(`${API_BASE}/models`).then(r => r.ok ? r.json() : null)
@@ -46,6 +74,13 @@ export function Settings() {
      .then(r => r.ok ? r.json() : null)
      .then(d => { if (d) setIconApps(d.apps || []); })
      .catch(() => {});
+
+    fetch(`${API_BASE}/status`)
+      .then(r => r.ok ? r.json() : null)
+      .then(d => { if (d) setPlatform(d.platform || ""); })
+      .catch(() => {});
+
+    loadViteStatus();
  }, []);

  const update = (key, value) => setForm(f => ({ ...f, [key]: value }));
@@ -167,7 +202,7 @@ export function Settings() {
  const displayQueue = isBranding ? brandQueue : selectedApps;

  return (
-    <div style={{ maxWidth: "720px" }}>
+    <div style={{ width: "100%" }}>
      <h2 style={{ margin: "0 0 1.25rem", fontSize: "1.1rem", color: "#eee" }}>Settings</h2>

      {/* Auto model routing */}
@@ -480,7 +515,42 @@ export function Settings() {
        </div>
      </div>

-      {/* Icon Branding */}
+      {/* Vite Dev Server */}
+      <div style={sectionStyle}>
+        <h3 style={{ margin: "0 0 0.5rem", fontSize: "0.95rem", color: "#bbb" }}>Vite Dev Server</h3>
+        <p style={{ margin: "0 0 1rem", fontSize: "0.75rem", color: "#555" }}>
+          Hot-reload server for editing the frontend directly (interface/web). Not
+          needed for normal use — this backend already serves the built UI.
+        </p>
+        <div style={{ display: "flex", alignItems: "center", gap: "0.75rem", flexWrap: "wrap" }}>
+          <button
+            onClick={toggleVite}
+            disabled={viteBusy || viteRunning === null}
+            style={{
+              padding: "0.6rem 1.25rem",
+              background: viteRunning ? "#2a1a1a" : "#152a15",
+              color: viteRunning ? "#ff8a80" : "#8aff8a",
+              border: "1px solid " + (viteRunning ? "#5a2a2a" : "#2a5a2a"),
+              borderRadius: "8px",
+              cursor: (viteBusy || viteRunning === null) ? "not-allowed" : "pointer",
+              fontSize: "0.88rem",
+              fontWeight: 500,
+              opacity: (viteBusy || viteRunning === null) ? 0.6 : 1,
+            }}
+          >
+            {viteBusy ? (viteRunning ? "Stopping…" : "Starting…")
+              : viteRunning ? "⏹ Stop Vite" : "▶ Start Vite"}
+          </button>
+          <span style={{ fontSize: "0.82rem", color: viteRunning ? "#8aff8a" : "#666" }}>
+            {viteRunning === null ? "Checking…" : viteRunning ? "Running at :5173" : "Stopped"}
+          </span>
+          {viteError && <span style={{ color: "#f44336", fontSize: "0.85rem" }}>{viteError}</span>}
+        </div>
+      </div>
+
+      {/* Icon Branding — Linux-only: composites icons onto XFCE/GTK app tiles,
+          which don't exist on Windows/macOS. */}
+      {platform === "linux" && (
      <div style={sectionStyle}>
        <h3 style={{ margin: "0 0 0.75rem", fontSize: "0.95rem", color: "#bbb" }}>App Icon Branding</h3>
        <p style={{ margin: "0 0 1rem", fontSize: "0.78rem", color: "#555" }}>
@@ -638,6 +708,7 @@ export function Settings() {
          )}
        </div>
      </div>
+      )}

      {/* Actions */}
      <div style={{ display: "flex", gap: "0.75rem", alignItems: "center" }}>
@@ -1,16 +1,21 @@
 # NexusOS launcher for Windows (native, no Vite).
 #
 # Single-process app: the backend on :8000 serves the built web UI itself, so
-# this starts only the memory service + backend, then opens the UI as an Edge
-# app window. The AI (Ollama) does NOT auto-start - turn it on from the UI's
-# Start AI button. Closing the app window stops the services.
+# this starts only the memory service + backend, opening a native app window
+# (bin\nexus_window.py, pywebview/WebView2) immediately alongside them with its
+# own live-updating loading screen. The AI (Ollama) does NOT auto-start - turn
+# it on from the UI's Start AI button. Closing the app window stops the
+# services.
 #
 #   Right-click -> Run with PowerShell   (or:  powershell -File launch_nexus.ps1)

 $ErrorActionPreference = "Stop"
+$ProgressPreference = "SilentlyContinue"  # Invoke-WebRequest's progress bar adds real latency for no benefit here
 $Root = $PSScriptRoot
 $Py = Join-Path $Root "Promethean\Scripts\python.exe"
 if (-not (Test-Path $Py)) { $Py = "python" }  # fall back to PATH
+$PyW = Join-Path $Root "Promethean\Scripts\pythonw.exe"
+if (-not (Test-Path $PyW)) { $PyW = "pythonw" }  # fall back to PATH

 $Runtime = Join-Path $Root "runtime"
 $Logs = Join-Path $Runtime "logs"
@@ -22,38 +27,41 @@ function Start-Svc($log, $argList) {
        -RedirectStandardOutput $log -RedirectStandardError "$log.err"
 }

+function Test-ServiceUp($url) {
+    try {
+        $r = Invoke-WebRequest -Uri $url -TimeoutSec 2 -UseBasicParsing
+        return $r.StatusCode -eq 200
+    } catch { return $false }
+}
+
 Write-Host "Starting NexusOS..." -ForegroundColor Cyan
-# Only start a service if its port is free. If it's already listening (a prior
-# launch, or started by hand) reuse it instead of spawning a duplicate that
-# fails to bind and stalls the readiness wait below.
+# Only reuse a port if something actually answers there. A bare port-listen
+# check isn't enough: a wedged process left over from a prior crashed launch
+# can be squatting the port without ever responding, and treating that as
+# "already running" skips starting a real service - the window then sits on
+# its loading screen for the full 40s with no way to tell why.
 $memory = $null
-if (-not (Get-NetTCPConnection -LocalPort 8001 -State Listen -ErrorAction SilentlyContinue)) {
+if (-not (Test-ServiceUp "http://127.0.0.1:8001/")) {
    $memory = Start-Svc (Join-Path $Runtime "memory.log")  @("-m","uvicorn","synapse.memory.service:app","--host","127.0.0.1","--port","8001")
 }
 $backend = $null
-if (-not (Get-NetTCPConnection -LocalPort 8000 -State Listen -ErrorAction SilentlyContinue)) {
+if (-not (Test-ServiceUp "http://127.0.0.1:8000/status")) {
    $backend = Start-Svc (Join-Path $Runtime "backend.log") @("-m","uvicorn","synapse.main:sio_app","--host","127.0.0.1","--port","8000")
 }

-# Wait for the backend to answer on :8000.
-$ok = $false
-foreach ($i in 1..30) {
-    try {
-        Invoke-WebRequest -UseBasicParsing -Uri "http://localhost:8000/status" -TimeoutSec 2 | Out-Null
-        $ok = $true; break
-    } catch { Start-Sleep -Seconds 1 }
-}
-if (-not $ok) {
-    Write-Host "Backend did not come up - check runtime\backend.log" -ForegroundColor Red
-}
-
 # Open the UI in a native window (pywebview / WebView2) - no browser, no Edge
-# profile cold-start. bin\nexus_window.py blocks until the window is closed and
-# falls back to the default browser if WebView2 is unavailable.
-$app = Start-Process -FilePath $Py -ArgumentList @((Join-Path $Root "bin\nexus_window.py")) `
+# profile cold-start - IMMEDIATELY, in parallel with memory/backend still
+# coming up. bin\nexus_window.py opens on its own loading page and polls
+# :8001 then :8000 itself, updating that page's status line live as each
+# comes up, so the window is on screen from the first instant instead of
+# this script blocking silently (in a hidden window) for up to 30s first and
+# only then showing anything. It blocks until the window is closed and falls
+# back to the default browser if WebView2 is unavailable.
+$WindowScript = Join-Path $Root "bin\nexus_window.py"
+$app = Start-Process -FilePath $PyW -ArgumentList "`"$WindowScript`"" `
    -WorkingDirectory $Root -PassThru

-Write-Host "NexusOS is running at http://localhost:8000  (AI is off - start it in the UI)" -ForegroundColor Green
+Write-Host "NexusOS window opened; services are starting at http://localhost:8000  (AI is off - start it in the UI)" -ForegroundColor Green
 # Block until the app window (or a service we started) exits, so this process
 # stays alive to hold and later stop the services. No blind Read-Host: the
 # shortcut runs hidden, where a prompt no one can answer would hang forever.
@@ -202,6 +202,14 @@ def check(svc: Service) -> bool:
    return False


+# Poll granularity for both waiters below. 1s steps used to mean every
+# start/stop paid up to a full second of dead latency per service on top of
+# however long the process actually took - three services in sequence could
+# lose several seconds to nothing but sleep(). 0.25s still amounts to one
+# cheap local HTTP HEAD every quarter second, not a busy-loop.
+_POLL_STEP = 0.25
+
+
 def wait_for_port(svc: Service, timeout: int = 30) -> bool:
    if http_ok(svc.url):
        # "READY", not "already running": `ncp start` launches memory and backend
@@ -210,8 +218,8 @@ def wait_for_port(svc: Service, timeout: int = 30) -> bool:
        # this same command started two seconds ago reads like a stale process.
        print(f"  {svc.label} READY (:{svc.port})")
        return True
-    for _ in range(timeout):
-        time.sleep(1)
+    for _ in range(int(timeout / _POLL_STEP)):
+        time.sleep(_POLL_STEP)
        if http_ok(svc.url):
            print(f"{svc.label} STARTED")
            return True
@@ -220,10 +228,10 @@ def wait_for_port(svc: Service, timeout: int = 30) -> bool:


 def wait_for_port_close(port: int, timeout: int = 15) -> bool:
-    for _ in range(timeout):
+    for _ in range(int(timeout / _POLL_STEP)):
        if not http_ok(f"http://localhost:{port}/"):
            return True
-        time.sleep(1)
+        time.sleep(_POLL_STEP)
    return False


@@ -353,24 +361,33 @@ def cmd_start(target) -> None:
    elif target in ("--ai", "-a"):
        start_ollama()
    elif target in (None, "", "all"):
-        # Bring the UI up first, then warm Ollama in the background — the model
-        # loads concurrently and into the first chat instead of blocking boot.
+        # Bring the UI up first, then kick off Ollama and (on Linux) Vite in the
+        # background without waiting on either — neither gates the app being
+        # usable. The backend already serves the built interface/web/dist at
+        # :8000 on both platforms (single-process design), and that's the URL
+        # nexus-app.sh/ncp web actually opens; nothing points a user at :5173.
+        # Vite only exists for whoever is hot-reload-editing the frontend, and
+        # they'll open :5173 themselves once it's ready - polling for it here
+        # just delayed "boot done" for a benefit nobody in the critical path
+        # gets. Skipped outright on Windows, where it's not part of the normal
+        # workflow at all and is the slow part of `ncp stop` to boot (npm's
+        # cmd.exe -> node -> esbuild tree doesn't die from a plain terminate()
+        # and falls through to the ~30s force-kill path). Still available on
+        # demand via `ncp start --frontend`.
        t0 = time.perf_counter()
        launch(SERVICES["memory"])
        launch(SERVICES["backend"])
        wait_for_port(SERVICES["memory"])
        wait_for_port(SERVICES["backend"])
        t_services = time.perf_counter()
-        launch(SERVICES["frontend"])
-        wait_for_port(SERVICES["frontend"])
-        t_frontend = time.perf_counter()
        start_ollama(background=True)
-        t_ollama = time.perf_counter()
+        if not WINDOWS:
+            launch(SERVICES["frontend"])
+        t_bg = time.perf_counter()
        print("\nBoot timing:")
-        print(f"  services (memory+backend) : {t_services - t0:5.1f}s")
-        print(f"  frontend (UI ready)       : {t_frontend - t_services:5.1f}s")
-        print(f"  ollama kickoff (bg warm)  : {t_ollama - t_frontend:5.1f}s")
-        print(f"  total to interactive      : {t_ollama - t0:5.1f}s")
+        print(f"  services (memory+backend)      : {t_services - t0:5.1f}s")
+        print(f"  ollama + frontend (bg kickoff) : {t_bg - t_services:5.1f}s")
+        print(f"  total to interactive           : {t_bg - t0:5.1f}s")
    else:
        show_help()

@@ -0,0 +1,109 @@
+"""Start/stop the Vite dev server from the web UI.
+
+Dev-only convenience: production serves the built interface/web/dist from this
+same backend, but hot-reload editing of the frontend needs Vite running
+separately. Shares management/ncp.py's PID file (runtime/pids/frontend.pid)
+and process patterns so `ncp status` / `ncp stop --frontend` see the same
+process regardless of which side started it.
+"""
+from __future__ import annotations
+
+import os
+import shutil
+import subprocess
+import urllib.request
+
+import psutil
+
+from .nexus_config import PROJECT_ROOT, RUNTIME_DIR
+
+FRONTEND_DIR = PROJECT_ROOT / "interface" / "web"
+PID_FILE = RUNTIME_DIR / "pids" / "frontend.pid"
+LOG_FILE = RUNTIME_DIR / "frontend.log"
+PORT = 5173
+
+
+def _npm() -> str | None:
+    return shutil.which("npm.cmd" if os.name == "nt" else "npm")
+
+
+def _read_pid() -> int | None:
+    try:
+        return int(PID_FILE.read_text().strip())
+    except (OSError, ValueError):
+        return None
+
+
+def _is_ours(pid: int) -> bool:
+    # On Windows, npm.cmd can't run as a CreateProcess image directly, so the
+    # OS wraps it as `cmd.exe /c ...npm.cmd run dev ...` - the ".cmd" splits
+    # "npm" from "run dev", so a literal "npm run dev" substring never matches
+    # the tracked PID's own cmdline. Loosen to "vite" (matches once the tree
+    # gets that far) or "npm"+"dev" both present (matches the wrapper hop too).
+    try:
+        cmd = " ".join(psutil.Process(pid).cmdline())
+    except Exception:
+        return False
+    return "vite" in cmd or ("npm" in cmd and "dev" in cmd)
+
+
+def _http_up() -> bool:
+    try:
+        with urllib.request.urlopen(f"http://127.0.0.1:{PORT}/", timeout=1.5) as r:
+            return r.status == 200
+    except Exception:
+        return False
+
+
+def is_running() -> bool:
+    """Mirrors ncp.py's own status check: PID liveness first, then whether the
+    port actually answers (covers Vite started by another process, or a lost
+    PID file) - not the stricter pattern match `stop()` uses before killing."""
+    pid = _read_pid()
+    if pid is not None and psutil.pid_exists(pid):
+        return True
+    return _http_up()
+
+
+def start() -> dict:
+    if is_running():
+        return {"status": "already_running"}
+    npm = _npm()
+    if not npm:
+        return {"status": "error", "detail": "npm not found - install Node.js"}
+
+    argv = [npm, "run", "dev", "--", "--host", "0.0.0.0"]
+    PID_FILE.parent.mkdir(parents=True, exist_ok=True)
+    LOG_FILE.write_text("")
+    # CREATE_NO_WINDOW (not DETACHED_PROCESS): a detached child has no console
+    # of its own, so Windows hands one to any console program it spawns in
+    # turn - npm.cmd -> node -> vite would each flash a window. CREATE_NO_WINDOW
+    # gives it a console that's never shown, inherited down the chain.
+    kwargs = ({"creationflags": subprocess.CREATE_NEW_PROCESS_GROUP
+               | getattr(subprocess, "CREATE_NO_WINDOW", 0)}
+              if os.name == "nt" else {"start_new_session": True})
+    with open(LOG_FILE, "ab") as log:
+        proc = subprocess.Popen(argv, cwd=str(FRONTEND_DIR), stdout=log,
+                                 stderr=subprocess.STDOUT, stdin=subprocess.DEVNULL,
+                                 **kwargs)
+    PID_FILE.write_text(str(proc.pid))
+    return {"status": "started", "pid": proc.pid}
+
+
+def stop() -> dict:
+    pid = _read_pid()
+    if pid is not None and psutil.pid_exists(pid) and _is_ours(pid):
+        try:
+            proc = psutil.Process(pid)
+            # npm.cmd -> node -> vite is a multi-hop tree; kill it depth-first
+            # so the parent doesn't outlive its children as an orphaned shell.
+            for child in proc.children(recursive=True):
+                try:
+                    child.terminate()
+                except Exception:
+                    pass
+            proc.terminate()
+        except Exception:
+            pass
+    PID_FILE.unlink(missing_ok=True)
+    return {"status": "stopped"}
@@ -10,6 +10,7 @@ from uuid import UUID

 import httpx
 import os as _os
+import platform as _platform
 from pathlib import Path
 from fastapi import FastAPI, HTTPException, Body
 from fastapi.middleware.cors import CORSMiddleware
@@ -19,6 +20,7 @@ from .nexus_config import settings, VERSION, DEFAULT_CHAT_MODEL
 from .chat import generate_chat_response, stream_chat_response, _synapse_trace
 from . import chat as _chat
 from .ollama_manager import initialize_ollama, initialize_ollama_async, get_ollama_manager
+from . import frontend_manager as _frontend_manager
 from .playbook_manager import PlaybookManager
 from . import tools as _tools

@@ -226,7 +228,7 @@ async def root():
                  if (ollama is not None and hasattr(ollama, "get_status")) else None)
    except Exception:
        status = None
-    return {"status": "online", "version": VERSION, "ollama": status}
+    return {"status": "online", "version": VERSION, "ollama": status, "platform": _platform.system().lower()}


 # -------------------------
@@ -788,6 +790,31 @@ async def ollama_stop_endpoint():
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

+# -------------------------
+# Vite Dev Server (Settings toggle) - dev-only convenience, not part of the
+# single-process production path. Runs via a thread so npm's own startup time
+# doesn't block the event loop, matching the Ollama start/stop pattern above.
+# -------------------------
+@app.get("/frontend/status")
+async def frontend_status_endpoint():
+    return {"running": await _asyncio.to_thread(_frontend_manager.is_running)}
+
+
+@app.post("/frontend/start")
+async def frontend_start_endpoint():
+    try:
+        return await _asyncio.to_thread(_frontend_manager.start)
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+
+@app.post("/frontend/stop")
+async def frontend_stop_endpoint():
+    try:
+        return await _asyncio.to_thread(_frontend_manager.stop)
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
 # -------------------------
 # Playbooks List (existing)
 # -------------------------