Back to Blog
Product UpdateDelivery HubInstall0.252

What's New on Cloud Nimbus — The Install Moment, Backup With Receipts, and One Operating Layer

GB
Glen Bradford
7 min read

Most Salesforce packages ambush you on day one. You click install, and suddenly there are notifications firing, emails queuing, jobs running, and a settings page full of toggles you didn't ask for. The first thing you do with a new tool is turn it down.

Delivery Hub 0.252 does the opposite. It installs silent — lights up only what's safe, leaves everything else off, and lets you opt into each capability when you're ready. That one change is the through-line for everything we shipped this wave, so here's the whole arc in one place.

Salesforce · Install PackageInstalling

Package verified

Delivery Hub 0.252 · Salesforce-native

Workflows ready

7 workflow types · 93 stages · 21+ persona views

Safe defaults on

Board metrics + activity logging — that's it

Silent by default

Notifications, email & AI off — zero inbox noise

Org Health receipt

Backup captured · count verified against your org

The install moment, on a loop — the same animation that now leads the homepage.

0.252
Latest Release
~3 min
To Install
2
Defaults On
0
Notifications

1. It Installs Silent

On a fresh install, exactly two things switch on: board metrics and activity logging. Both are read-only, both are useful from minute one, and neither sends anyone a message. Notifications, email, AI assistance, and every automation signal start off. You turn them on, one at a time, when you understand what each one does — not because a package decided for you.

This sounds like a small thing. It isn't. The number-one reason internal tools get abandoned in the first week is that they were loud before they were useful. A silent install means the first impression is the board, not the inbox.

2. Backup That Hands You a Receipt

Most Salesforce backup tools give you a confirmation email. Org Health gives you a receipt — the captured record count for every object, verified against your org's own count, every run. If the two numbers don't match, you find out before you need the backup, not after.

Every run also produces a utilization report (which fields are actually populated) and a list of deprecation candidates (fields sitting at zero or near-zero fill). Run it on a schedule and the snapshots become a record of how your schema is decaying over time — the kind of thing nobody measures until a migration forces them to. The open-source engine is live and running on real production data today; the managed service is on the way.

See it at /org-health or drop a backup folder into the backup explorer to browse integrity, fill rates, and decay trends in the browser.

3. The Client Portal Runs the Real Gantt Engine

The customer-facing portal timeline used to render through a separate demo renderer — close to the in-org Gantt, but not the same code. It now mounts the exact production engine the in-org timeline and the embedded Lightning component use, with the same NOW / NEXT / PLANNED priority buckets. Your clients see the same picture your team does, from the same engine, with no Salesforce license required.

One engine, three surfaces: the in-org record page, the embedded component, and the public portal. When we improve the Gantt, all three get it at once.

4. A Public API That Behaves Under Load

The public API picked up the three things every integration eventually needs and most never get: rate-limiting with a real Retry-After so clients back off instead of hammering, idempotency on the toggle endpoints so a retried request doesn't double-fire, and a pagination envelope on list endpoints so large result sets page cleanly. The portal client now understands all three — it surfaces a friendly “retry in N seconds” instead of a raw error, and can drive next-page UI without re-parsing.

5. The Changelog Caught Up

The public changelog had been stranded at an April release while the package itself shipped all the way to 0.252. It's backfilled now — themed per release, not per-PR, with the highlights called out: the zero-noise install defaults, the API hardening, the delivery-health signals, and the feature-catalog scaffold. Read the whole run at /changelog.

The Bigger Picture: One Install, a Whole Operating Layer

Step back and the pattern is clear. That three-minute install isn't dropping an app into Salesforce — it's dropping in an operating layer. The same install becomes the place your team tracks work, the portal your clients watch progress in, the catalog you procure productized workflows through, and the backup that proves your data is safe.

  • The platformWork tracking, invoicing, e-signature, cross-org sync, and AI — free and Salesforce-native.
  • For your clientsA real-time portal with no Salesforce license and no status emails.
  • Productized workEleven vertical workflows you can procure, all running on the same platform.
  • Peace of mindBackup with an integrity receipt and schema-decay analytics.

None of these is individually the headline. The pattern underneath them is: every piece is both a feature you use and a primitive the next piece is built on. And it all starts with an install that doesn't make a sound.

Try It

The managed package is free. Install it into any Salesforce org with standard Setup permissions — it's an in-place upgrade from any prior version, and it stays quiet until you decide otherwise.

Previous roundup: the spring write-up on right-click menus, live charts, and the sync chain is here.