agentui-wc 0.1.60 → 0.1.62

package/AGENTS.md

@@ -1954,7 +1954,7 @@ dist/
 ├── agentui.css          # Combined CSS (90 KB)
 ├── agentui.d.ts         # TypeScript definitions
 ├── routes/              # Route Bundles (Modern)
-│   ├── shell.js         # App Shell (Navbar, Theme, Bus) - 5 KB
+│   ├── shell.js         # App Shell (Navbar, Theme, Bus) - 13 KB raw / 5 KB gzip
 │   ├── home.js          # Home route components
 │   ├── chunk-*.js       # Shared dependencies (deduplicated)
 │   ├── route-deps.json  # Auto-generated page dependencies
@@ -2080,7 +2080,7 @@ For Modern performance, use the **App Shell** pattern. Load the critical shell f
 
 | Bundle | Content | Size (gzip) |
 |--------|---------|-------------|
-| `shell.js` | Navbar, Theme, EventBus, Toasts | ~5 KB |
+| `shell.js` | Navbar, Theme, EventBus, Toasts | ~5 KB gzip |
 | `home.js` | Landing page components | ~2 KB |
 | `forms.js` | (Example) Input, Button, Checkbox | ~7 KB |
 
@@ -2092,7 +2092,7 @@ For Modern performance, use the **App Shell** pattern. Load the critical shell f
 ---
 
 ## Performance Metrics
-- **Bundle size:** 173 KB full / ~20 KB initial shell (gzipped: 60 KB / ~20 KB)
+- **Bundle size:** 173 KB full / ~20 KB initial shell (gzipped: 60 KB total / ~20 KB initial)
 - **Lighthouse:** 100% Performance, 100% Accessibility, 100% SEO, 100% Best Practices
 - **EventBus:** 111M ops/sec (LightBus emitSync)
 - **Virtual List:** Handles 10K+ items at 60fps

package/README.md

@@ -75,7 +75,7 @@ AgentUI is built on four ideas:
 | **Bundle** | 60KB gzipped (JS+CSS), all 50 components | Tree-shake to ~20KB, then add deps |
 | **Performance** | Lighthouse 100/100/100/100 ([verify](https://pagespeed.web.dev/analysis?url=https://giuseppescottolavina.github.io/AgentUI/demo/)) | Varies |
 | **Dependencies** | Zero. Forever. | npm audit anxiety |
-| **Tests** | 1300+ tests, 0 failures, isolated runner | Coverage varies |
+| **Tests** | 1300+ tests (unit + E2E), 0 unit failures | Coverage varies |
 | **Memory** | Managed listeners (AbortController), zero leaks verified | Framework-dependent |
 | **Security** | Every component XSS-audited, CSP-compatible, no `eval()` | Trust your deps |
 | **Stability** | W3C Web Components — no migration treadmill | React 16→17→18→19... |
@@ -117,13 +117,13 @@ AgentUI is in **alpha** (v0.1.x). Here's an honest look at where we are:
 - ARIA roles, keyboard navigation, and 48px minimum touch targets
 
 **Performance**
-- 60KB total gzipped (48KB JS + 13KB CSS) — includes all 50 components
+- 60KB total gzipped (47KB JS + 13KB CSS) — includes all 50 components
 - Route-based lazy loading with Speculation Rules API prefetching
 - Lighthouse 100/100/100/100 on demo site ([verify it yourself](https://pagespeed.web.dev/analysis?url=https://giuseppescottolavina.github.io/AgentUI/demo/))
 - 500 component DOM instantiation in <8ms, 500 updates in <3ms
 
 **Engineering**
-- 1300+ isolated tests, zero failures, across 99 test files
+- 1300+ tests (868 unit + E2E), zero unit failures, across 99 test files
 - Every component XSS-audited — `escapeHTML()` on all user-facing content
 - Zero runtime dependencies — zero supply chain risk
 - Managed event listeners via AbortController — verified zero memory leaks
@@ -137,15 +137,33 @@ AgentUI is in **alpha** (v0.1.x). Here's an honest look at where we are:
 
 > 💡 **Long-term maintainability by design** — AgentUI is built on W3C Web Components, a browser standard — not a framework. There's no migration from v16 to v17. No breaking Hooks rewrite. No Composition API pivot. `querySelector` has worked the same since 1998 and will work the same in 2030. Zero dependencies means zero `npm audit` surprises, zero abandoned transitive packages, and zero "waiting for library X to support framework Y."
 
-### What's missing (and we know it)
-- Larger community testing and feedback
-- More complex data patterns (autocomplete, tree view, rich text)
-- Accessibility audit by a11y experts
-- Agent workflow benchmarks vs. other frameworks
-- Formal agent feedback studies
+### Roadmap — Designed, Ready to Build
 
-**This list is not a roadmap — it's an invitation.**
-These are problems we'd love to solve *together*.
+The foundation (50 components, agent API, build system) is solid. Here's what's next — each feature is architecturally planned and waiting for hands to build it:
+
+**Phase 1 — Enterprise Essentials**
+
+| Feature | What & Why |
+|---------|------------|
+| 🚦 **Router v2** | History API, `/users/:id` params, nested layouts, route guards. The current hash router works for docs sites — enterprise PWAs need real URLs and auth middleware. |
+| 🧠 **Data Query Layer** (`au-query`) | Declarative data fetching with automatic caching, request deduplication, and global interceptors (JWT). Think TanStack Query, but as a web component with `describe()`. |
+| 🌍 **i18n** | Native `<au-t key="...">` with reactive language switching. LightBus already provides the event backbone — the wiring is straightforward. |
+
+**Phase 2 — Scale & Power**
+
+| Feature | What & Why |
+|---------|------------|
+| ⚡ **Virtual DataTable** | `au-datatable` already handles sorting/filtering. Adding a `virtual` mode (leveraging the existing `au-virtual-list` engine) enables 10K+ row datasets. |
+| 🏗️ **App Shell wrapper** | `au-layout` + `au-drawer` + `au-bottom-nav` already compose into a full app shell (the demo uses this). A convenience `<au-app-shell>` would make this one-liner instead of manual composition. |
+| ⌨️ **Command Palette & Hotkeys** | Global `Ctrl+K` search, configurable shortcuts with conflict resolution. Essential for power-user PWAs competing with native apps. |
+
+**Phase 3 — Agent Intelligence**
+
+| Feature | What & Why |
+|---------|------------|
+| 🧪 **Auto-Test Generator** | `describe()` already tells agents what every component can do. A CLI that reads schemas and generates baseline tests would eliminate boilerplate and let teams focus on business logic. |
+
+> 💡 **Want to accelerate this?** Every feature above has a clear spec and fits the existing architecture. [Start a discussion](https://github.com/GiuseppeScottoLavina/AgentUI/discussions) or open a PR — contributions move the needle fast on a project this focused.
 
 
 ---

package/dist/AGENTS.md

@@ -1954,7 +1954,7 @@ dist/
 ├── agentui.css          # Combined CSS (90 KB)
 ├── agentui.d.ts         # TypeScript definitions
 ├── routes/              # Route Bundles (Modern)
-│   ├── shell.js         # App Shell (Navbar, Theme, Bus) - 5 KB
+│   ├── shell.js         # App Shell (Navbar, Theme, Bus) - 13 KB raw / 5 KB gzip
 │   ├── home.js          # Home route components
 │   ├── chunk-*.js       # Shared dependencies (deduplicated)
 │   ├── route-deps.json  # Auto-generated page dependencies
@@ -2080,7 +2080,7 @@ For Modern performance, use the **App Shell** pattern. Load the critical shell f
 
 | Bundle | Content | Size (gzip) |
 |--------|---------|-------------|
-| `shell.js` | Navbar, Theme, EventBus, Toasts | ~5 KB |
+| `shell.js` | Navbar, Theme, EventBus, Toasts | ~5 KB gzip |
 | `home.js` | Landing page components | ~2 KB |
 | `forms.js` | (Example) Input, Button, Checkbox | ~7 KB |
 
@@ -2092,7 +2092,7 @@ For Modern performance, use the **App Shell** pattern. Load the critical shell f
 ---
 
 ## Performance Metrics
-- **Bundle size:** 173 KB full / ~20 KB initial shell (gzipped: 60 KB / ~20 KB)
+- **Bundle size:** 173 KB full / ~20 KB initial shell (gzipped: 60 KB total / ~20 KB initial)
 - **Lighthouse:** 100% Performance, 100% Accessibility, 100% SEO, 100% Best Practices
 - **EventBus:** 111M ops/sec (LightBus emitSync)
 - **Virtual List:** Handles 10K+ items at 60fps

package/dist/README.md

@@ -75,7 +75,7 @@ AgentUI is built on four ideas:
 | **Bundle** | 60KB gzipped (JS+CSS), all 50 components | Tree-shake to ~20KB, then add deps |
 | **Performance** | Lighthouse 100/100/100/100 ([verify](https://pagespeed.web.dev/analysis?url=https://giuseppescottolavina.github.io/AgentUI/demo/)) | Varies |
 | **Dependencies** | Zero. Forever. | npm audit anxiety |
-| **Tests** | 1300+ tests, 0 failures, isolated runner | Coverage varies |
+| **Tests** | 1300+ tests (unit + E2E), 0 unit failures | Coverage varies |
 | **Memory** | Managed listeners (AbortController), zero leaks verified | Framework-dependent |
 | **Security** | Every component XSS-audited, CSP-compatible, no `eval()` | Trust your deps |
 | **Stability** | W3C Web Components — no migration treadmill | React 16→17→18→19... |
@@ -117,13 +117,13 @@ AgentUI is in **alpha** (v0.1.x). Here's an honest look at where we are:
 - ARIA roles, keyboard navigation, and 48px minimum touch targets
 
 **Performance**
-- 60KB total gzipped (48KB JS + 13KB CSS) — includes all 50 components
+- 60KB total gzipped (47KB JS + 13KB CSS) — includes all 50 components
 - Route-based lazy loading with Speculation Rules API prefetching
 - Lighthouse 100/100/100/100 on demo site ([verify it yourself](https://pagespeed.web.dev/analysis?url=https://giuseppescottolavina.github.io/AgentUI/demo/))
 - 500 component DOM instantiation in <8ms, 500 updates in <3ms
 
 **Engineering**
-- 1300+ isolated tests, zero failures, across 99 test files
+- 1300+ tests (868 unit + E2E), zero unit failures, across 99 test files
 - Every component XSS-audited — `escapeHTML()` on all user-facing content
 - Zero runtime dependencies — zero supply chain risk
 - Managed event listeners via AbortController — verified zero memory leaks
@@ -137,15 +137,33 @@ AgentUI is in **alpha** (v0.1.x). Here's an honest look at where we are:
 
 > 💡 **Long-term maintainability by design** — AgentUI is built on W3C Web Components, a browser standard — not a framework. There's no migration from v16 to v17. No breaking Hooks rewrite. No Composition API pivot. `querySelector` has worked the same since 1998 and will work the same in 2030. Zero dependencies means zero `npm audit` surprises, zero abandoned transitive packages, and zero "waiting for library X to support framework Y."
 
-### What's missing (and we know it)
-- Larger community testing and feedback
-- More complex data patterns (autocomplete, tree view, rich text)
-- Accessibility audit by a11y experts
-- Agent workflow benchmarks vs. other frameworks
-- Formal agent feedback studies
+### Roadmap — Designed, Ready to Build
 
-**This list is not a roadmap — it's an invitation.**
-These are problems we'd love to solve *together*.
+The foundation (50 components, agent API, build system) is solid. Here's what's next — each feature is architecturally planned and waiting for hands to build it:
+
+**Phase 1 — Enterprise Essentials**
+
+| Feature | What & Why |
+|---------|------------|
+| 🚦 **Router v2** | History API, `/users/:id` params, nested layouts, route guards. The current hash router works for docs sites — enterprise PWAs need real URLs and auth middleware. |
+| 🧠 **Data Query Layer** (`au-query`) | Declarative data fetching with automatic caching, request deduplication, and global interceptors (JWT). Think TanStack Query, but as a web component with `describe()`. |
+| 🌍 **i18n** | Native `<au-t key="...">` with reactive language switching. LightBus already provides the event backbone — the wiring is straightforward. |
+
+**Phase 2 — Scale & Power**
+
+| Feature | What & Why |
+|---------|------------|
+| ⚡ **Virtual DataTable** | `au-datatable` already handles sorting/filtering. Adding a `virtual` mode (leveraging the existing `au-virtual-list` engine) enables 10K+ row datasets. |
+| 🏗️ **App Shell wrapper** | `au-layout` + `au-drawer` + `au-bottom-nav` already compose into a full app shell (the demo uses this). A convenience `<au-app-shell>` would make this one-liner instead of manual composition. |
+| ⌨️ **Command Palette & Hotkeys** | Global `Ctrl+K` search, configurable shortcuts with conflict resolution. Essential for power-user PWAs competing with native apps. |
+
+**Phase 3 — Agent Intelligence**
+
+| Feature | What & Why |
+|---------|------------|
+| 🧪 **Auto-Test Generator** | `describe()` already tells agents what every component can do. A CLI that reads schemas and generates baseline tests would eliminate boilerplate and let teams focus on business logic. |
+
+> 💡 **Want to accelerate this?** Every feature above has a clear spec and fits the existing architecture. [Start a discussion](https://github.com/GiuseppeScottoLavina/AgentUI/discussions) or open a PR — contributions move the needle fast on a project this focused.
 
 
 ---

package/dist/component-schema.json

@@ -4,7 +4,7 @@
   "title": "AgentUI Component Schema",
   "description": "Machine-readable schema for AI agent component discovery",
   "version": "0.1.4",
-  "generatedAt": "2026-02-08T07:13:24.377Z",
+  "generatedAt": "2026-02-08T12:16:37.048Z",
   "componentCount": 50,
   "components": {
     "au-card": {

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "agentui-wc",
-    "version": "0.1.60",
+    "version": "0.1.62",
     "type": "module",
     "description": "AI-First Web Components Framework - 50 Material Design 3 components optimized for AI agents",
     "main": "dist/agentui.esm.js",