⭐️ Features

Data Transformation

#

Filtering

#

$filter supports structured conditions with field, op, and value.

Supported operators:

• eq, neq
• gt, gte, lt, lte
• contains
• exists
• regeq, regneq

Child $filter overrides parent $filter when both are present.

Sanitization

#

• $pick: keep only selected fields.
• $omit: remove selected fields.

Visibility Control

#

Two directives withhold output, and they differ in reach.

$emit controls endpoint emission at the node it appears on.

• $emit: ["list"] emits only collection endpoints.
• $emit: ["ids"] emits only per-item endpoints.
• $emit: ["list", "ids"] emits both.
• $emit: [] emits neither.

$emit does not reach the node’s sub-paths, and a sub-path with no $emit of its own emits everything. So this publishes the whole dataset one level down, despite the empty $emit:

{
	"staff": {
		"$emit": [],
		"by-name": {}
	}
}

$skip: true leaves the node and everything beneath it ungenerated. A skipped node is never descended into, so no sub-path can publish through it — including one added later by someone who did not notice the $skip, and including a descendant that sets $skip: false.

{
	"staff": {
		"$skip": true,
		"by-name": {}
	}
}

Nothing under /staff is written, and none of it appears in the discovery index. Use $emit: [] to drop one node’s own endpoints, and $skip when a subtree must stay ungenerated whatever is added to it.

$skip is named for what it does to the build. Nothing is generated, which is not the same as generating something and protecting it — static hosting has no access control, and this tool offers none.

$private, $emit_list, $emit_id and $emit_items are not accepted. $private was described in earlier documentation but never worked; $skip is the directive that does what it described.

Aggregation

#

$aggregate bundles multiple endpoints or collections into one endpoint payload.

Template Expansion

#

Template keys such as ${year} can be expanded by:

• $values for static lists.
• $derive for values extracted from loaded source data, with an optional type (string, int) so derived values can be compared against non-string fields.

Build and Runtime Experience

#

• Incremental builds via file hash cache.
• Automatic TypeScript .d.ts generation.
• Automatic media asset copy and path rewrite for referenced local files.

Platform Integration

#

• Automatic _headers generation for Cloudflare Pages and Netlify.
• Automatic vercel.json generation for Vercel CORS/header behavior.

Serializer Flexibility

#

• JSON output for static API endpoints.
• TypeScript/JavaScript module output for frontend build pipelines.
• SQLite output for offline/local query scenarios.
• Per-serializer minify option.