---
title: "Changelog"
description: "Release notes for meirdick/auto-search-chips."
---

> Documentation Index
> Fetch the complete documentation index at: https://auto-search-chips.mdick85.workers.dev/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Changelog

## 0.1.0 (unreleased)

First public version.

- Rules engine: enum values, relation names, number comparisons, dates, booleans, sort words, `field: value` syntax and quoted phrases.
- AI engines: Cloudflare System One (Clairic) and chat models through `laravel/ai`. Custom engines with `SearchChips::extend()`.
- Two-phase parse: rules chips at once, then the AI refine. Suggested chips for uncertain AI readings.
- `searchChips(ChipSchema $chips)` model definition, zero-config discovery, and `search-chips:schema --stub`.
- `SearchChips::for()`, the `autoSearch()` scope, `fromRequest()`, `toProps()` and `applyChips()`.
- Components for Livewire, Blade and Alpine, React, Vue and Svelte, on one headless TypeScript core. The box holds the typed text. Chips, keyword text and sort show under it.
- Filter panel for Livewire and React. It shows the chips as controls, in sync with the search box.
- Value synonyms, keyword search over related names, and `onlyWhen()` row conditions.
- Name check: with Clairic, a bare name is an author chip only when that author exists. Zero-config keyword search reads related names. The key and the timestamps are not sent to an AI engine.
- Value matching: with Clairic, a Clef decision model maps a word such as "movies" to the allowed value `video`.
- Laravel Scout: keyword text goes through Scout for `Searchable` models.
- Laravel Scout: chips go to the engine as filters when it can apply them all (`collection`, `database`, Meilisearch, Typesense). Pages and totals then come from the engine. Config `scout.filters`.
- `search-chips:scout {alias}`: prints the index settings a model needs.
- `meirdick/laravel-cf-workersai` 1.3 or later as the System One transport when installed.
- JSON routes with auth, a rate limit, per-model authorization and a base query.
- Schema cache, AI parse cache and `SearchChips::fake()`.

Source: https://auto-search-chips.mdick85.workers.dev/docs/changelog/index.mdx
