Cloud deployments, sites, applications and data pipelines that "just work"

Deploy your applications, backend stack and data pipelines on a blazing fast cloud that just works.

Status: Preview. Join to start shipping application and data infrastructure with the skippr-cloud CLI.

Skippr services are designed to be simple and powerful. They behave as you'd expect, without you needing engineering faff to make services consistent, auto-tune, or bill sanely.

Skippr's design emphasises good taste, predictability, and efficiency. We always choose auto-tuning, consistency, and correctness by default. Decades of building developer tooling and platforms for engineers in VC-funded startups and enterprise have taught us one thing above all else: a simple system is a good system.

Deploying Skippr Cloud services like serverless functions, Sites CDN, CI/CD runners, or complex data pipelines should feel effortless — elegant, joyful even. Scaling, billing, and data consistency should all be predictable. Skippr is an opinionated cloud, easy to reason about... because you and your LLMs have enough to think about.

It just works

• Tables look like the DynamoDB-shaped store you already know — partition keys, conditions, transactions — but behave as you first expected: every read is strongly consistent (always!), and a GSI updates in the same commit as the item. No stale-index surprise. And full SQL support! See Tables.
• Auth: App user pools cover users for your application, with simple tenancy, JWT, and OTP built in. Not hundreds or thousands of product config matrix knobs. Just fast, secure authentication for people who use your product. See Auth.
• Cloud User Directory: Sign in to Skippr Cloud with email, manage people on the account, and attach policies so a function or CI identity can call Cloud services. See Cloud User Directory.
• ELT discovers schemas, evolves them as sources change, and loads any supported source to any destination. The runner is multi-threaded and auto-tunes. You do not set worker counts, deal with offsets, or batch knobs. Just correct, exactly-once data, every time... even when your source data has breaking schema changes. It's ELT that, like everything in Skippr, "just works". See ELT.
• Internals auto-tune. Pools, channels, and batching adapt under fixed safety ceilings. You configure the service (table name, pipeline, schedule), and crack on with developing your application.

Start in a minute

curl -fsSL https://install.skippr.io/install.sh | sh
skippr-cloud login --email you@example.com

Paste the one-time code from email. That is sign-in. Windows: irm https://install.skippr.io/install.ps1 | iex.

Full Cloud install: CLI and SDKs. Self-hosted engine: skipprd install. Cloud walkthrough: Get started.

Deploy a backend

Create a table, put an item, read it back. Keys and typed attributes are familiar; consistency is not a flag you have to remember.

skippr-cloud tables create-table --table-name notes --input - <<JSON
{
  "attributeDefinitions": [{ "attributeName": "pk", "attributeType": "S" }],
  "keySchema": [{ "attributeName": "pk", "keyType": "HASH" }]
}
JSON

skippr-cloud tables put-item --table-name notes --input - <<JSON
{
  "item": {
    "pk": { "S": "note-1" },
    "body": { "S": "ship it" }
  }
}
JSON

skippr-cloud tables get-item --table-name notes --input - <<JSON
{ "key": { "pk": { "S": "note-1" } } }
JSON

Point HTTP routes at a function with Gateway. Store files in Objects. Queue background work with Queue.

Deploy a site

skippr-cloud sites create-site --name acme

You get a Skippr hostname. Upload a deployment, promote it when it is ready, bind a custom domain after DNS. Previews are first-class. Same-origin APIs can go to Gateway. See Sites.

Build pipelines and analytics

Self-hosted ingest is skipprd discover / skipprd sync. Data Engineer workflows are sde init, sde connect, and sde model. Cloud skippr-cloud is not those binaries.

sde init orders
sde connect
skipprd discover --pipeline orders
skipprd sync --pipeline orders --once
sde model --pipeline orders

Managed Cloud ELT is skippr-cloud elt on cloud.skippr.io/elt. Engine docs: skippr.io. Data Engineer: data-engineer.skippr.io.

Pricing that just works

When Cloud billing ships, you pay four physical meters: vCPU time, memory time, bytes stored, network bytes. Not request packs, not RCU/WCU, not a matrix of seat SKUs. Same idea as the services: easy to reason about.

Skippr Cloud is Preview and does not publish a public price list yet. There is no live Cloud invoice path today. Contact for Preview access, an ELT demo, or a usage discussion. Detail: Pricing.

Where to go next

You want	Go here
Sign in and run a Cloud command	Get started
Account people, policies, and keys	Cloud User Directory
App user pools	Auth
Consistent document store	Tables
Website and previews	Sites
HTTP or WebSocket API	Gateway
Discover, load, model	ELT
Terraform / CDKTF	Providers
Wire contracts	API reference
Talk to us	Contact