datasette-ts

TypeScript-first Datasette-style explorer for SQLite, with local serve and Cloudflare Workers deploy.

Install

npm install -g datasette-ts
# or
npx datasette-ts --help

Quickstart (local)

datasette-ts serve ./my.db --port 8001

Open http://127.0.0.1:8001.

Inspect a database

datasette-ts inspect ./my.db --inspect-file inspect-data.json

Deploy to Cloudflare

First, set up Alchemy (docs):

npx alchemy configure
npx alchemy login

Then deploy:

# Deploy with name derived from filename
datasette-ts deploy cloudflare ./my.db

# Deploy with explicit name
datasette-ts deploy cloudflare ./my.db --name my-app

# Deploy with a specific Alchemy profile
datasette-ts deploy cloudflare ./my.db --name my-app --profile prod

# Deploy with metadata (robots settings, cache settings, etc.)
datasette-ts deploy cloudflare ./my.db --metadata datasette.yml

This creates a Cloudflare Worker and D1 database with your data.

CLI help

datasette-ts --help
datasette-ts serve --help
datasette-ts inspect --help
datasette-ts deploy cloudflare --help

HTTP caching

Defaults to Cache-Control: max-age=5 on all responses. You can override this default using a setting, or per-request using the _ttl query string parameter:

# Set the default max-age for all responses (in seconds)
datasette-ts serve ./my.db --setting default_cache_ttl 60

# Override cache behavior for a single request
curl "http://127.0.0.1:8001/?_ttl=0"
curl "http://127.0.0.1:8001/?_ttl=120"

You can also set default_cache_ttl via metadata:

settings:
  default_cache_ttl: 60

• default_cache_ttl of 0 disables caching by sending Cache-Control: no-store.
• _ttl=0 disables caching for that request; _ttl=<seconds> sets max-age.

Hashed URLs

Hashed URLs include a content hash in the database path (e.g. /mydb-abc1234/table), enabling long-lived immutable caching. When enabled, responses include Cache-Control: public, max-age=31536000, immutable.

Enable via metadata:

settings:
  hash_urls: true

Or per-request with ?_hash=1. Cloudflare Workers enable hashed URLs by default (set DATASETTE_HASH_URLS=0 to disable).

Robots.txt

Serve a robots.txt that blocks crawlers by default. Configure it via metadata:

{
	"plugins": {
		"datasette-block-robots": {
			"allow_only_index": true
		}
	}
}

Options:

• allow_only_index: allow indexing the homepage while blocking each database path.
• disallow: custom list of paths to disallow, e.g. ["/mydb/mytable"].
• literal: full robots.txt contents (overrides all other settings).

The /robots.txt endpoint returns the generated file with text/plain content.

Build

npm run build

Build outputs:

• dist/cli.js (Node CLI entrypoint)
• dist/worker.js (Cloudflare Worker entrypoint)

The deploy command prefers dist/worker.js when present, so the published package ships a ready-to-bundle worker.

Publish

npm version <patch|minor|major>
npm run build
npm pack
npm publish

Status

• Works: local serve, inspect data export, Cloudflare Workers deploy via Alchemy.
• Not yet: config files, plugins, write APIs.