Atlas Admin Directory
Find and open admin, hosting, and monitoring links for every Atlas site — fast.
| Tags |
|---|
No sites match
Try a different search term or clear your filters to see the full directory.
Tools
General-purpose links for internal tools and vendor portals — not tied to any specific site or server.
No tools match
Try a different search term.
Reports
Quick operational reports across the directory — starting with domains that need renewal attention.
Expiring domains
Sorted soonest-first. Domains not on auto-renew need a person to actually renew them before the date passes.
| Site | Domain | Expires | Days left | Auto-renew | Registrar |
|---|
Nothing expiring in this window
Try a wider window above to see further out.
Work & billing
Defaults to Unbilled work, all dates, oldest first, so older outstanding items never scroll out of sight.
| Work date | Customer | Site | Description | Staff | Work order | Billing status | Amount | Billed date |
|---|
Admin
User accounts, inventory data, and the system-wide message banner.
Users
No email/reset-link flow exists in this build — set a new password directly and share it with the user out of band. Select a user (or Add User) to open their edit card.
Inventory data
Replaces the whole directory with the sites in the imported file (local to this browser; doesn't touch the shared Atlas database).
The sample is a CSV matching scripts/csv-to-json.js's columns — convert it with that script before using "Import inventory" above, which expects JSON. A direct CSV import isn't built yet.
Discards this browser's imported inventory and reloads the bundled default. Asks for confirmation and is recorded below.
System message
Shown prominently to every signed-in user, on every page, until removed here — no expiry, no per-user dismiss.
Database snapshot
Stored outside the web root; credentials and paths never reach the browser. A note is required for every manual snapshot so its purpose is on record.
Available snapshots
Retention policy
Automatic cleanup only ever runs when explicitly enabled below — it starts disabled. Proposed defaults: keep all successful snapshots 7 days, one daily for 30 days, one weekly for 12 weeks, one monthly for 12 months; remove failed attempts after 7 days. The newest successful snapshot, and any snapshot held for an active import/restore/investigation, can never be removed.
Cleanup dry run
Read-only — shows exactly what the current policy would remove. Never deletes anything by itself. Run this first; the real cleanup below is disabled until you have.
Run cleanup
Restore from snapshot
Required before restoring
- A fresh automatic before-restore snapshot of the current database will be created first — this is not optional and cannot be skipped.
- Maintenance mode will be turned on for the duration of the attempt.
- A post-restore schema/connectivity/count health check will run before maintenance mode is ever turned back off.
- If verification fails, maintenance mode stays on until an admin reviews it.
Operational Assets — import preview
Read-only dry run for the 182-record operational-assets backfill (domain/DNS/mail records with no matching website). Recomputes and validates the reconciliation against current source data; never writes to the database. Full JSON/CSV artifacts are saved outside the web root.
Operational Assets — apply approved import
Run the preview above first — approval is tied to that exact preview (by fingerprint); if the source data changes, re-run the preview before applying.
| Domain | Category | Capabilities | Notes / evidence |
|---|
Website Data Migration — 435 existing websites
Read-only dry run for the canonical database foundation (site_domains, site_hosting, site_mail, site_links, site_source_metadata). Recomputes and validates 435 = 435 = 435 (Websites sheet / inventory.json / sites) and 0 failed site-key resolutions against current source data; never writes to the database.
Field groups to apply
Records to apply
| Domain | Customer | Domain Relationship | Notes |
|---|
Deferred / review-only (never written as canonical, regardless of selection)
Tag catalog
Tags are governed: an admin defines each one once (name, description, an approved colour and icon), and admins and operators then apply those tags to websites and operational assets. Tags are independent of website type, Runs WordPress, customer link, cleanup outcome, QB comparison and flagged notes. A tag that has ever been used is retired, never deleted, so its history stays; only a tag that was never applied can be deleted.
Merge tags
| Tag | Status | Websites | Assets | Created / last changed | Actions |
|---|
Tag history
| When | Who | Action | Tag | Record | Change |
|---|
Customer Cleanup
Nothing here is auto-linked. Accepting a candidate or creating a customer always sets a provisional link — confirm or remove it later without losing the original raw value or review history. Link Status is the authority on what still needs confirming.
QB Ref / Note (Optional) — every field in that column is optional and may hold a QuickBooks customer name, a QuickBooks customer ID/reference, and/or a short reconciliation note. It is kept separate from the original source value, the proposed candidate, the QB comparison, and the customer link. The current QuickBooks export contains no customer IDs, so none are filled in — a reference is only ever what someone types.
Outcome vs Customer link status are different things: Outcome records how a link was originally made (e.g. "Provisional accepted") and never changes again once set, even after the link itself is later confirmed — it is a history label, not a live status. To find links that still need confirming, filter by Customer link status = Provisional, not by Outcome.
| Site / Asset | Original source value | Proposed candidate | Relationship | Customer link | QB Comparison | Outcome | Assignee | QB Ref / Note (Optional) | Actions |
|---|
Recent activity
Admin actions logged from this page — inventory resets, database snapshots, customer changes, and other admin actions.
| Actor | Action | Details | Timestamp |
|---|
Customer Maintenance
Search, view, and edit customer records. Site assignment stays on each site's detail panel.
Operational Assets
Domain, DNS, and mail records that aren't tracked websites yet. Propose a customer link or classification change below; an admin confirms or rejects it.
Servers
The canonical server/droplet inventory. A server's own fields live here once, not copied into every site it hosts; sites link to a server from their own Recovery / Hosting panel.
Migrations
Track planned server moves for sites in the Nomad migration queue. Add a site here from its own detail page to start tracking its move.
| Customer | Site | Current server | Status | Phase | Blocker | Next action |
|---|