Do these work with MDX?

Yes — the article shell styles a content slot rather than parsing anything, so MDX, a CMS payload or plain children all render the same. The prose styles are scoped to that slot rather than applied globally.

What structured data should an article carry?

`Article` with a headline, an author and both dates. It is cheap, it is accurate if you generate it from frontmatter, and it is the difference between a bare result and one with a byline.

Are these blog and content page layouts free to use?

10 of the 14 are free to copy and use commercially; 4 need a Pro licence, and each one says so on its own page before you spend any time on it.

How do I add one to my project?

Install it with the CLI: `npx hoverlab add blog-index-page` writes the component and its dependencies into your project. Every artifact here also has a copy button for the raw source, and an "Open in sandbox" link if you would rather try it first.

Why does this page mix components and whole pages?

Because the phrase does. The same search covers 8 composed pages and 6 templates — one of them is the right size for what you are building, and which one depends on how much of the screen you still have to write.

Name	Type	Required	Description
project_id	string	Required	The project to list deployments for. Passed in the path.
status	string	Optional	Filter by state: building, ready or failed.
limit	integer	Optional	Page size, 1–100. Defaults to 20.
cursor	string	Optional	Opaque cursor from a previous page’s next_cursor.

Install

$ npm install ratchet

MIT · v2.4.0 · 2.1M downloads a month

Schema migrations that refuse to break production

A migration runner that reads your application code before it drops a column, and blocks itself when something still needs it. Plain SQL, one static binary, no DSL.

$ratchet up

ratchet 2.4.0 — schema migrations that refuse to go backwards

✓ 0001_initial.sql applied 2.4s

✓ 0002_add_ledger.sql applied 0.8s

✓ 0003_split_settlement.sql applied 14.1s

⚠ 0004_drop_legacy.sql BLOCKED

0004 drops column `orders.legacy_ref`, which 2 queries still read.

Found in: reports/monthly.sql:41, api/orders.go:212

Run with --force to override, or delete the readers first.

3 applied, 1 blocked, 0 failed — database unchanged for 0004.

Read the docs Star on GitHub

The whole API is one object and two files

A config, and migrations that are plain SQL. If you already know how to write an ALTER TABLE you know how to use this.

Migrations are SQL files — nothing to learn, nothing to unlearn
The guard is the only unusual concept, and it is four lines
Irreversible migrations are declared, not inferred

1import { Ratchet } from 'ratchet'2 3// A migration is a file, not a class. The up is required; the down is4// optional and, if you leave it out, the migration is declared5// irreversible and the tool will say so rather than guess.6export const ratchet = new Ratchet({7  dir: './migrations',8  url: process.env.DATABASE_URL,9 10  // The whole point of the library. Before applying anything that drops11  // or narrows a column, parse the queries in these paths and refuse if12  // one of them still reads it.13  guard: {14    scan: ['api/**/*.go', 'reports/**/*.sql'],15    onConflict: 'block',16  },17})18 19await ratchet.up()

Why this rather than the one you already have

Five differences. The first one is the reason the project exists; the others are consequences of keeping it small.

It reads your code before it drops a column

The guard parses the queries in the paths you name and blocks a destructive migration while anything still reads the column. Most outages caused by a migration are this exact shape.

Plain SQL, no DSL

Your database’s own documentation is the reference. Nothing to learn and nothing that stops working when Postgres adds a feature.

Irreversible is a thing you declare

A migration with no down is fine — lying about it is not. Say so in a comment and the plan output says so too.

No lock held while it thinks

Planning happens before the transaction opens, so a slow guard scan never holds a lock on a production table.

One binary, no runtime

6MB, static, no Node or Python needed in the image that runs it. The library and the CLI are the same code.

18.4k

GitHub stars

And 312 open issues, honestly

2.1M

Downloads a month

npm, last 30 days

184

Contributors

41 with commit access

MIT

Licence

Since 2021, and it will not change

Funding, since you are about to depend on it

Two maintainers are paid part-time; the rest is volunteered

Northwind funds one day a week and Contoso funds two, both on rolling twelve-month agreements, and neither gets any say over the roadmap — that is written into both. If both stopped tomorrow the project would keep going more slowly rather than stopping, because no single contributor holds more than 19% of commits. This paragraph is here because it is the question a team should ask before taking a dependency, and most projects do not answer it.

£61k: Funded in 2025; Two corporate sponsors, 340 individuals
3 days: Paid maintainer time a week; Across two people
19%: Largest share of commits; No single point of failure
0: Sponsor influence on the roadmap; Written into both agreements

Where to go next

The documentation is the place to start. The Discord is faster on weekdays, and every design decision is argued in a public RFC.

GitHub

The code, the issues and the RFC process. Every design decision was argued in public and the threads are still there.

18.4k stars · 312 open issues

Discord

About 4,000 people. Two maintainers are in it daily; nobody is paid to be, so answers are faster on weekdays.

4,100 members

Documentation

Guides, a full CLI reference and thirteen recipes for migrations that are genuinely hard to get right.

Versioned back to 1.0

Changelog

Every release since 1.0, with the breaking changes called out and a codemod where one was possible.

Latest: 2.4.0

v4.2 — no agent, no sidecar

Webhooks that arrive, or tell you why not

Relay takes delivery, retries, signature verification and the dead-letter queue off your critical path. One endpoint in, one verified payload out.

$npm i @relay/sdk

Read the docs See pricing

Delivering for

Northwind
Contoso
Initech
Globex
Lumon
Vandelay

4.1 billion events delivered last month, at 99.99% first-attempt success.

The whole integration

Two files. There is no third one where the complexity was hiding.

Signature verification is one call, and it throws rather than returning false
Retries, backoff and the dead-letter queue are server-side — nothing to schedule
Typed payloads generated from your own event schema

1import { verify } from '@relay/sdk'2 3export async function POST(req: Request) {4  // Throws on a bad signature or a replayed timestamp, so there is5  // no boolean to forget to check.6  const event = await verify(req, process.env.RELAY_SECRET!)7 8  switch (event.type) {9    case 'invoice.paid':10      await fulfil(event.data.invoiceId)11      break12    case 'invoice.failed':13      await notify(event.data.customerId)14      break15  }16 17  // 200 means "stored". Relay retries anything else for 72 hours.18  return new Response(null, { status: 200 })19}

What you stop maintaining

Each of these is a thing teams build in-house, ship, and then own forever.

Delivery

The retry loop you were going to write on Friday

Twelve attempts over 72 hours with exponential backoff, resumed across your deploys rather than lost with the process that scheduled them. Anything that never lands goes to a dead-letter queue you can replay from a dashboard or a single API call.

Retries survive your restarts
Replay one event or ten thousand
Per-endpoint circuit breaking

Verification

Signatures, timestamps and replay windows, once

Constant-time comparison, a configurable tolerance window, and rejection of anything reused. The failure mode of hand-rolled verification is not an outage — it is silently accepting forged events for a year, which is why this is the section to read twice.

Constant-time compare
Replay rejection by default

Visibility

Every attempt, with the response body attached

The request, the response, the status code and the latency, for every attempt of every event, kept for 30 days. When a customer says the webhook never arrived, this is the difference between an answer and an afternoon.

Full request and response bodies
30-day retention
Filter by endpoint, type or status

Works with what you already run

Direct integrations, not a webhook you have to adapt at both ends.

Stripe
Payments
Available
Verified passthrough of every Stripe event type.
GitHub
Source
Available
Repository, org and app events, with installation scoping.
Shopify
Commerce
Available
Order, fulfilment and inventory topics.
Slack
Alerting
Available
Post failures into a channel with the payload attached.
AWS EventBridge
Infrastructure
Beta
Forward into a bus with the original signature preserved.
Kafka
Infrastructure
Planned
Produce to a topic with at-least-once semantics.

What it costs at your volume

Move the slider. There is no plan to pick, and the number below is the number on the invoice.

Monthly volume2,000,000 events

100K50M

0–1M at $0.08 per 1K: $80.00
1M–10M at $0.05 per 1K: $50.00

Estimated monthly total$130.00

Get an API key

The awkward questions

Infrastructure buyers ask about exits before they ask about features. Fair enough.

What happens if Relay goes down?: Your senders keep retrying into us and we keep the backlog; nothing is dropped for the length of your retry window. If we are down longer than that, the failover endpoint documented in the SDK forwards raw to your origin unverified, and you fall back to your own handling.
Can I self-host it?: Yes, on the enterprise plan, as a container with a Postgres and a Redis. It is the same image we run. The dashboard and the replay API come with it; the only thing that does not is our on-call.
How do I get my data out?: A single export endpoint returns every event, attempt and response body as newline-delimited JSON, and it is available on every plan including the free one. No ticket, no notice period, no export fee.
Is the SDK required?: No. It is 40 lines around a documented HTTP contract and an HMAC. The verification algorithm is written out in the docs so you can implement it in a language we do not ship for — several customers have.
What counts as an event?: One inbound payload, regardless of how many times we retry it or how many endpoints you fan it out to. Retries are our problem, so charging you for them would be charging you for our failures.
Do you have an SLA?: 99.99% on ingest, credited automatically against the next invoice rather than on request. The status page below is the same one our own alerting reads, and its history is not editable after the fact.

Blog and content page layouts

Pages8

Templates6

Questions about blog and content page layouts

Related collections