docs(readme): document workspace-root routing, Tool Optimization (@mux), auto-update

- README: add "Different Tools for Different Folders" (per-folder workspace-root
  routing) and "Let Your AI Curate Its Own Toolset" (built-in Tool Optimization
  driven by the @mux trigger; reads silent, writes approved, space_id targeting).
  Note auto-install-on-launch updates in How It Works.
- Capture spec: add scenes for the Workspaces and Tool Optimization pages so the
  pipeline produces workspaces.png + tool-optimization.png.
- docs/screenshots/README: document the two new screenshots.

Screenshots for the two new surfaces are populated by running the capture
command (see docs/screenshots/README.md) on a machine with a display.

Signed-off-by: Mohammod Al Amin Ashik <maa.ashik00@gmail.com>

Author: its-mash

README.md

@@ -79,7 +79,7 @@ Lightweight and cross-platform — built in Rust with Tauri 2, McpMux uses minim
 
 **3.** Done. Every tool from every server is available in every client, right now.
 
-McpMux routes calls to the right server, refreshes OAuth tokens automatically, and keeps credentials encrypted in your OS keychain — you never think about it again.
+McpMux routes calls to the right server, refreshes OAuth tokens automatically, and keeps credentials encrypted in your OS keychain — you never think about it again. It also keeps **itself** current: new versions download and install on launch by default (toggle in Settings), so a restart is all it takes.
 
 ---
 
@@ -107,6 +107,12 @@ Create isolated Spaces — each with their own servers, credentials, and permiss
 
 ![Workspaces — switch context instantly from the sidebar](docs/screenshots/space-switcher.png)
 
+### Different Tools for Different Folders
+
+Your AI client tells McpMux which folder it's working in (its MCP *root*). McpMux uses that to **route each workspace to its own toolset** — open your backend repo and the AI sees your database and deploy tools; open a docs folder and it sees only search and filesystem. Map a folder once in the **Workspaces** tab (or let the AI do it — see below) and every future session from that exact path resolves automatically. Matching is per-folder and exact, so nothing leaks across projects.
+
+![Workspaces — route each folder to its own FeatureSet](docs/screenshots/workspaces.png)
+
 ### Control What Each Client Can Do
 
 Not every AI client should have the same power. Create Feature Sets — permission bundles that control exactly which tools, prompts, and resources a client can access. Build a "Read Only" set for cautious workflows, a "React Development" set with just GitHub and Filesystem, or a "Full Stack Dev" set with everything. Assign them per-client so each tool only goes where you want it.
@@ -119,6 +125,22 @@ Cursor, VS Code, Windsurf, Claude Code — see every AI client connected to your
 
 ![Client Management — per-client permissions and effective features](docs/screenshots/client-detail.png)
 
+### Let Your AI Curate Its Own Toolset
+
+Hand an assistant a hundred tools and it burns tokens and reaches for the wrong one. McpMux ships a built-in **Tool Optimization** capability so the AI can keep *itself* lean — straight from chat, no config files.
+
+Start a request with **`@mux`** and the assistant can:
+
+- **Discover** what's available — `mcpmux_list_spaces`, `mcpmux_list_all_tools`, `mcpmux_search_tools`
+- **Compose** a focused FeatureSet of just the tools it needs — `mcpmux_manage_feature_set`
+- **Pin** the current folder to that set so it sticks — `mcpmux_bind_current_workspace`
+
+Reads are silent; anything that changes your setup pops a **one-click approval dialog that names the exact Space** — the AI proposes, you decide. The `@mux` trigger keeps these requests cleanly separated from your real work, and every operation can target a specific Space by id.
+
+> *"@mux build a minimal toolset for this Next.js repo and pin it to this folder."*
+
+![Tool Optimization — the AI composes and pins its own toolset, gated by your approval](docs/screenshots/tool-optimization.png)
+
 ---
 
 ## Security

docs/screenshots/README.md

@@ -26,6 +26,8 @@ pnpm exec wdio run tests/e2e/wdio.conf.ts --spec tests/e2e/specs/capture-screens
 | `discover.png` | Discover | content area | Server registry (mock bundle with 14 servers) |
 | `spaces.png` | Spaces | content area | Workspace cards (default + 5 additional) |
 | `featuresets.png` | FeatureSets | content area | Permission bundles |
+| `workspaces.png` | Workspaces | content area | Per-folder routing — map a folder (MCP root) to a FeatureSet/Space |
+| `tool-optimization.png` | Built-in Servers | content area | Tool Optimization shelf: meta-tool list + `@mux` trigger tip |
 | `clients.png` | Clients | full window | OAuth consent modal over connected clients (Cursor, VS Code, Windsurf, Claude Code) |
 | `settings.png` | Settings | content area | App settings page |
 

tests/e2e/specs/capture-screenshots.manual.ts

@@ -528,6 +528,37 @@ describe('Screenshot Capture', function () {
     await saveScreenshot('featuresets');
   });
 
+  it('captures Workspaces (per-folder routing)', async () => {
+    const nav = await byTestId('nav-workspaces');
+    await safeClick(nav);
+    await browser.pause(2000);
+    // Open the "map a folder" form so the binding UI (root + FeatureSet picker)
+    // is visible even without a live client reporting roots.
+    try {
+      const toggle = await byTestId('workspace-binding-create-toggle');
+      await safeClick(toggle);
+      await browser.pause(800);
+    } catch {
+      // form may already be open or the page empty — capture whatever renders
+    }
+    await saveScreenshot('workspaces');
+  });
+
+  it('captures Tool Optimization (built-in self-management)', async () => {
+    const nav = await byTestId('nav-builtin-servers');
+    await safeClick(nav);
+    await browser.pause(2000);
+    // Select the Tool Optimization capability to reveal its tool list + @mux tip.
+    try {
+      const card = await byTestId('builtin-server-tool-optimization');
+      await safeClick(card);
+      await browser.pause(800);
+    } catch {
+      // fall back to the shelf overview
+    }
+    await saveScreenshot('tool-optimization');
+  });
+
   it('captures Connected Clients with approval modal', async () => {
     const nav = await byTestId('nav-clients');
     await safeClick(nav);