PLAN: the KVM tab, as simply as it can actually be built

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

PLAN.md

@@ -553,37 +553,39 @@ something long in one, switch away and back, refresh — it's all still there.
   item names the heavier pieces (multi-select and bulk actions, folder download, resumable
   uploads, a gallery, search) — worth doing the two together, since they are the same tab.
 
-- **A KVM tab** — asked for 2026-09-24, not started. A tab that shows the **IP-KVM's own web
+- **A KVM tab** — asked for 2026-09-24, not started. **Dead simple is the brief**: paste one
-  page** inside Ham Lab, so a machine's screen, keyboard and mouse are one click away. **This
+  HTTP link, the tab shows whatever UI that KVM serves, and the keyboard goes to it. **This
   is KVM the keyboard-video-mouse switch, not KVM the hypervisor** — the Virtualization item
-  above means the other one, and the two must not be confused when either is built.
+  above means the other one.
-  - **The device**: a **Sipeed NanoKVM**, attached to `grid`, at **`10.0.0.217`** (MAC
+  - **The device**: a **Sipeed NanoKVM** attached to `grid` at **`10.0.0.217`** (MAC
-    `48:da:35:6f:d0:31`), serving a web UI on port 80. The only one in the lab so far, so the
+    `48:da:35:6f:d0:31`), web UI on port 80. The only one so far, so store a list, not one URL.
-    tab must be built for a list of them, not hard-wired to one.
+  - **Framing works.** Probed 2026-09-24: it sends **no `X-Frame-Options` and no
-  - **Framing works.** Probed 2026-09-24: the NanoKVM sends **no `X-Frame-Options` and no
+    `frame-ancestors`**, so it does not refuse to be embedded. No fallback needed.
-    `frame-ancestors`**, so unlike most of its competitors it does not refuse to be embedded.
+  - **Proxy it through the daemon — that is what makes the simple version possible.** The
-    An iframe is a real option and the fallback of "just link out" is not needed.
+    KVM is **HTTP only**, and Ham Lab is served over TLS, so a bare `http://…` iframe is
-  - **Mixed content is the actual blocker.** The NanoKVM is **HTTP only** — there is no
+    blocked as mixed content and the "just paste a link" idea dies there. Serving it from Ham
-    HTTPS listener at all — and Ham Lab is served over TLS, so a browser will refuse the
+    Lab's own origin (`/api/kvm/<id>/…`, on the `proxyNode` machinery, which already carries
-    iframe outright. Something must give it HTTPS. Two routes:
+    WebSockets — exactly what the video and input streams are) fixes three things at once:
-    - **Proxy it through the daemon**, which is the better one. Ham Lab already
+    no mixed content, the KVM sits behind **Ham Lab's** login rather than a second one
-      reverse-proxies nodes over WebSockets (`proxyNode`), which is exactly what the KVM's
+    (`/api/vm/info` answers 401, so it has its own, which the hub can hold once), and the
-      video and input streams need, and it puts the KVM behind **Ham Lab's own login** rather
+    iframe becomes **same-origin**, which matters for the keyboard below.
-      than a second one. `/api/vm/info` answers **401**, so the device has its own auth; the
+  - **The keyboard needs less than it sounds.** A parent page *cannot* capture keys and
-      hub could hold that credential once instead of prompting per person.
+    forward them into a **cross-origin** iframe — the browser forbids synthetic events across
-    - **Or an NPM proxy host** (`scripts/npm-proxy-host.sh`), which is two minutes of work —
+    origins — but it does not need to: an iframe with focus receives keystrokes directly, and
-      but note what it means: `*.bigham.dev` is a wildcard to the public IP, so that would put
+    the NanoKVM's own page already handles them. So "capture and give it to the iframe" is
-      **BIOS-level keyboard and screen access to `grid` on the public internet** behind
+    really: click-to-focus, keep focus there, and stop the console's own shortcuts from
-      whatever the KVM's own login is worth. If that route is taken it wants an NPM access
+    eating keys while it is focused. What is genuinely worth adding is **fullscreen** plus
-      list or a LAN/WireGuard-only binding, decided deliberately.
+    `navigator.keyboard.lock()`, which is what stops the *browser* swallowing Escape, Tab and
-  - **Where the entry lives is still the design decision.** A KVM earns its keep when the
+    Ctrl-W so they reach the guest — note it is Chromium-only, so Firefox and iOS Safari get
+    the plain version. Proxying makes the iframe same-origin anyway, so real forwarding stays
+    available if some KVM ever needs it.
+  - **Where the list lives is the one real design decision.** A KVM earns its keep when the
     machine is *down* — no OS, a bad kernel, a BIOS screen — which is when that host's Ham Lab
-    cannot answer. So the list has to be **hub-side** and must work for machines running no
+    cannot answer. So it is **hub-side** and must work for machines running no Ham Lab at all.
-    Ham Lab at all. A `kvm_url` inside each node's own `config.json` would fail in precisely
+    A `kvm_url` in each node's own `config.json` would fail in exactly the case it is for.
-    the case it exists for.
+  - **Do not give it an NPM host** without deciding to: `*.bigham.dev` is a wildcard to the
-  - Worth settling too: whether the KVM shows on its machine's page or on one page listing
+    public IP, so that would put BIOS-level control of `grid` on the internet behind only the
-    every KVM; and whether a machine with no Ham Lab deserves a first-class entry on the
+    KVM's own login. Behind Ham Lab's proxy it inherits Ham Lab's auth instead.
-    Machines page, which today only knows hosts that answer.
 
 ## Open questions