feat(server): register the routes, the slot, the leg and the boot hooks

The entry point becomes real: five mount prefixes, the admin.users.detail
extension slot, the shard push catalog, the town-crier announce leg and both
lifecycle hooks. module.json declares all of it and the loader checks the
declaration against what register() actually registers, in both directions.

The URLs are byte-identical to the ones core served before the extraction. That
is the whole point of moving the code and not the paths: the shipped Android app
calls POST /api/v1/admin/shard/kick and the Discord bot reads
/api/v1/public/shard/*, and neither knows a module answers now.

Require order is load-bearing and the requires are inside register() because of
it. Every ported file reaches core through ./core, whose members resolve ctx
when called -- but a router does `const express = core.express` at ITS file
scope, which runs the moment it is required. Hoisting these to the top of the
file breaks the module with an error about ctx being missing, from a file that
never mentions it.

boot.js takes the eight UO call sites out of core's server.js. One behavioural
change, deliberate: uoLinkSocket.start() and the sidecar health probe used to
run AFTER the listener bound and now run before it, because onBoot does. start()
returns as soon as the reconnecting client is armed, but the probe is a real
HTTP call, so it is fired and NOT awaited -- an unreachable sidecar must not
hold the site closed. Reporting that the bridge is down is diagnostics; being up
is not a precondition for serving a page.

router/rateLimits.js builds the market limiter through ctx.middleware.rateLimit,
core's factory. The policy is the module's -- only the module knows what its
endpoints cost -- and the plumbing is core's, so there is one express-rate-limit
in the process and one place a breach is logged.

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

server/router/admin/shardAtlas.controller.js

@@ -0,0 +1,117 @@
+// ── Admin · Spawn atlas ────────────────────────────────────────────────────
+//
+// Operating the atlas import: where the ServUO tree is, whether it has drifted
+// from what is loaded, and the approve/reject decision for a refresh that would
+// remove a facet (docs/website/SPAWN_ATLAS.md).
+//
+// The policy lives in the model. This controller does three things and no more:
+// it validates input, it maps a refresh RESULT onto an HTTP status, and it
+// records the action in the admin activity log.
+//
+// **A refresh result is not an exception.** `shardAtlas.refresh()` reports
+// `unavailable` / `failed` / `needsReview` rather than throwing, because the boot
+// path must never be stopped by a bad tree. That contract is preserved here: an
+// unreadable mount is a 200 carrying `status: 'unavailable'`, not a 500. The
+// admin needs to be told what is wrong with their path, and a 500 says only
+// "something broke".
+
+const atlas = require('../../model/shardAtlas/shardAtlas.model')
+const { activity } = require('../../core')
+
+const log = require('../../core').logger('admin-shard-atlas')
+
+// GET /admin/shard/atlas — what is loaded, what the tree looks like, what is
+// staged. Unlike the public /atlas/meta route this DOES carry the filesystem
+// path and the drift flag: that is the whole point of the panel.
+async function getStatus(req, res) {
+  try {
+    return res.json(await atlas.status())
+  } catch (err) {
+    log.error('getStatus', err)
+    return res.status(500).json({ message: 'Internal Server Error' })
+  }
+}
+
+// POST /admin/shard/atlas/import — apply a map change without a restart.
+//
+// `force` reimports even when the source hashes match what is loaded (the escape
+// hatch for "the database is wrong but the tree is not"). Facet loss is still
+// staged rather than applied — approving is a separate, explicit act.
+async function importAtlas(req, res) {
+  try {
+    const force = !!req.body?.force
+    const result = await atlas.refresh({ force })
+    await activity.log({
+      req,
+      action: 'shard.atlas.import',
+      detail: { force, status: result.status, counts: result.counts ?? null },
+    })
+    return res.json(result)
+  } catch (err) {
+    log.error('importAtlas', err)
+    return res.status(500).json({ message: 'Internal Server Error' })
+  }
+}
+
+// POST /admin/shard/atlas/approve — apply a staged refresh, facet loss and all.
+//
+// Re-parses the tree rather than applying something captured at boot: only the
+// DECISION was stored, so what lands matches the tree as it is now. If the
+// operator has since fixed a half-copied mount, the approved import is simply
+// the corrected one — which is the desired outcome, not a surprise.
+async function approve(req, res) {
+  try {
+    const result = await atlas.approvePending()
+    await activity.log({
+      req,
+      action: 'shard.atlas.approve',
+      detail: { status: result.status, removed: result.removedFacets ?? null },
+    })
+    return res.json(result)
+  } catch (err) {
+    log.error('approveAtlas', err)
+    return res.status(500).json({ message: 'Internal Server Error' })
+  }
+}
+
+// POST /admin/shard/atlas/reject — keep the current atlas and remember the
+// decision against those exact source hashes, so a declined refresh does not
+// re-prompt on every restart. Changing the tree asks again.
+async function reject(req, res) {
+  try {
+    const result = await atlas.rejectPending()
+    if (result.status === 'none') {
+      return res.status(404).json({ message: 'No refresh is awaiting review.' })
+    }
+    await activity.log({ req, action: 'shard.atlas.reject', detail: {} })
+    return res.json(result)
+  } catch (err) {
+    log.error('rejectAtlas', err)
+    return res.status(500).json({ message: 'Internal Server Error' })
+  }
+}
+
+// PUT /admin/shard/atlas/path — point the atlas at a different ServUO tree.
+//
+// Persisted as a setting, which wins over the SERVUO_PATH env default so an
+// operator can move the mount without a redeploy. Blank clears it, which turns
+// the atlas off (boot skips, the loaded atlas keeps serving) — that is a
+// legitimate thing to want, so it is allowed rather than validated away.
+//
+// Deliberately does NOT import as a side effect: changing where the atlas reads
+// from and reloading it are separate decisions, and an operator fixing a typo
+// should not have a multi-thousand-row replace happen under them. The response
+// carries the refreshed status so the panel can offer the import immediately.
+async function setPath(req, res) {
+  try {
+    const value = String(req.body?.path ?? '').trim()
+    await atlas.setServuoPath(value, req.user?.id ?? null)
+    await activity.log({ req, action: 'shard.atlas.path', detail: { path: value } })
+    return res.json(await atlas.status())
+  } catch (err) {
+    log.error('setAtlasPath', err)
+    return res.status(500).json({ message: 'Internal Server Error' })
+  }
+}
+
+module.exports = { getStatus, importAtlas, approve, reject, setPath }