Skip to content

CHANGELOG.md

Latest commit

 

History

History
39 lines (30 loc) · 2.5 KB

CHANGELOG.md

File metadata and controls

39 lines (30 loc) · 2.5 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[0.1.0] — 2026-03-07

Initial release.

Added

Hooks

  • useTokenStream(source) — consumes any AsyncIterable<string> or ReadableStream<Uint8Array> and accumulates text, tracking the full streaming lifecycle (idle → streaming → complete | error). Exposes text, isStreaming, status, error, abort, and reset.
  • useMessageStream(parts) — bridges Vercel AI SDK UIMessage.parts to text, reasoning, hasActiveToolCall, hasReasoning, and sourceUrls.
  • useDebouncedStreaming(isStreaming, debounceMs?) — debounces the streaming boolean to prevent flicker between tokens.
  • useAISDKStatus(chatStatus) — maps Vercel AI SDK useChat status strings to StreamStatus.

Components

  • <StreamingText> — renders text that grows token-by-token with an optional blinking cursor (built-in or custom node). Renders as any HTML element via as prop. Fully unstyled.
  • <TypingIndicator> — animated "AI is thinking" indicator with dots, pulse, and bar variants. Inherits color from parent. Fully accessible.
  • <PartialRender> — renderer-prop component with error boundary for graceful partial-content rendering during streaming.
  • <StreamGuard> — status-driven render slots (idle, streaming, complete, error) for type-safe lifecycle switching.

Adapters (streaming-ui-primitives/adapters)

  • fromFetchSSE(response, options?) — converts a raw fetch Response to AsyncIterable<string>. Modes: auto, text, sse-text, sse-json, vercel-ai.
  • fromOpenAIChatStream(stream) — wraps OpenAI SDK chat.completions.stream().
  • fromOpenAICompletionStream(stream) — wraps OpenAI SDK legacy completions.create({ stream: true }).
  • fromAnthropicStream(stream) — wraps Anthropic SDK messages.stream().
  • partsToText(parts, options?) — extracts plain text from UIMessage.parts.
  • hasActiveToolCall(parts) — returns true if any tool invocation is pending.
  • fromUseChatStatus(status) — maps Vercel AI SDK useChat status to StreamStatus.

Infrastructure

  • Dual ESM + CJS build with full .d.ts declarations via vite-plugin-dts.
  • GitHub Actions CI (unit tests, typecheck, E2E via Playwright against built Storybook).
  • GitHub Actions OIDC-compliant npm publish on version tags.
  • MIT license.