addis_ai_sdk

Documentation

• Added Addis AI artwork and a polished centered header to the GitHub and pub.dev package README.
• Bundled the optimized source artwork in the repository so the package branding remains maintainable.

Documentation

• Redesigned the pub.dev README with a feature matrix, secure installation guidance, current API examples, platform support, migration notes, and links.
• Reworked the Flutter example to load credentials through --dart-define instead of suggesting a source-code API key.
• Updated the voice example to use the current voice catalog and durable clip generation API.

Fixed

• Restored the canonical OSI-approved MIT license text so automated package scoring and license detection work correctly.

Added

• OpenAI-style chat completions, SSE streaming, system prompts, personas, and automatic function-calling loops.
• Current voice generation, estimates, wallet usage, clip management, voice catalog, and preview APIs.
• Speech-to-text and translation resources.
• Retry-aware HTTP transport, API key/JWT authentication, idempotency keys, request options, and normalized production error types.
• Compatibility entrypoint and migration aliases for existing applications.
• topP and topK generation controls, reusable multimodal file references, realtime readiness events, PCM MIME envelopes, and upload validation.

Changed

• Aligned public behavior and endpoints with the official Addis AI JavaScript and Python SDKs at version 0.2.0.
• Removed production debug logging and corrected legacy streaming request data.

Fixed

• corrected license text and tightened dependencies

Added

• Initial release of Addis AI Flutter SDK
• Chat generation, streaming, TTS, realtime, attachments support
• Example app with voice/chat UI and logo icon

• TODO: Describe initial release.