gui: Danger Zone in the Updates tab — forget-all / factory reset

[mrjeeves]

What

Adds an in-app way to reset a node, so recovering a wedged install doesn't mean hand-deleting files under ~/.myownmesh. A Danger Zone card at the bottom of Settings → Updates with two graduated, two-click-armed actions:

• Forget all meshes — leave and purge every network (rosters + signed governance state), keeping the device identity.
• Factory reset — wipe the entire state directory (identity, config, all networks); the device becomes brand-new to every peer.

Why the reboot

Both actions reboot the whole stack when they fire. The daemon is the real datastore, so a reset that only deletes files gets undone by in-memory caches re-persisting on the next write — the classic "I deleted it but it came back." So each command wipes on disk, then the daemon exits; the GUI relaunches the app, and a fresh daemon comes up on clean state (via the existing ensure_daemon_running path, or the supervising service).

Changes

Daemon control protocol (crates/myownmesh/src/control.rs)

• Request::ForgetAllNetworks — bulk NetworkRemove{purge:true} over every network in the registry.
• Request::FactoryReset — quiesce each network (so nothing re-persists mid-wipe), then remove_dir_all the state dir.
• schedule_daemon_exit() — both exit ~400 ms after the response flushes, so a fresh daemon reloads clean.

GUI (gui/src-tauri, gui/src)

• mesh_forget_all_networks / mesh_factory_reset / restart_app Tauri commands (restart_app waits out the daemon's exit, then app.restart()).
• meshClient.forgetAllNetworksAndRestart / factoryResetAndRestart.
• The Danger Zone UI in UpdatesSection.svelte (confirm-armed buttons, always shown so it's reachable even if update status fails to load).

Testing

Daemon + Tauri crates: cargo build / clippy --all-targets -D warnings / fmt clean (pinned 1.88.0). Frontend: svelte-check clean (0 errors).

⚠️ The wipe and backend are compile-verified, but the reboot lifecycle (daemon self-exit → app relaunch → fresh daemon) can't be exercised in CI — it wants a smoke test on a real desktop install. The AllMyStuff side (its node supervises the daemon and also owns the ownership file + the fleet-leave level) will follow as a separate PR.

🤖 Generated with Claude Code

https://claude.ai/code/session_018x1RV6ppUdMDTHxScVo26L

---

Generated by Claude Code