Packages

altar_ai

0.1.0
0.1.0

Protocol-based AI adapter foundation for Elixir. Provides unified abstractions for gemini_ex, claude_agent_sdk, and codex_sdk with automatic fallback support, runtime capability detection, and built-in telemetry.

HexDocs

Current section

Files

Jump to
Readme
1 Version
2 Dependencies
2 Dependants
Files
Activity
altar_ai CHANGELOG.md
Raw

CHANGELOG.md 

# Changelog



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



The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),

and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).



## [0.1.0] - 2025-01-15



### Added



- Initial release of Altar.AI

- Protocol-based AI adapter architecture:

  - `Altar.AI.Generator` - Text generation and streaming

  - `Altar.AI.Embedder` - Embedding generation interface

  - `Altar.AI.Classifier` - Text classification interface

  - `Altar.AI.CodeGenerator` - Code generation and explanation

- Adapters:

  - `Altar.AI.Adapters.Gemini` - Google Gemini API adapter

  - `Altar.AI.Adapters.Claude` - Anthropic Claude API adapter

  - `Altar.AI.Adapters.Codex` - OpenAI Codex/GPT API adapter

  - `Altar.AI.Adapters.Composite` - Multi-provider fallback chain

  - `Altar.AI.Adapters.Mock` - Testing and development mock adapter

  - `Altar.AI.Adapters.Fallback` - Heuristic-based fallback adapter

- Utilities:

  - `Altar.AI.Error` - Unified error handling

  - `Altar.AI.Response` - Response normalization helpers

  - `Altar.AI.Classification` - Classification result type

  - `Altar.AI.CodeResult` - Code generation result type

  - `Altar.AI.Telemetry` - Built-in telemetry instrumentation

  - `Altar.AI.Capabilities` - Runtime capability detection

- Main API:

  - `Altar.AI.generate/2` - Text generation

  - `Altar.AI.stream/2` - Streaming text generation

  - `Altar.AI.embed/2` - Embedding generation

  - `Altar.AI.batch_embed/2` - Batch embedding generation

  - `Altar.AI.classify/3` - Text classification

  - `Altar.AI.generate_code/2` - Code generation

  - `Altar.AI.explain_code/2` - Code explanation

  - `Altar.AI.capabilities/1` - Query adapter capabilities

  - `Altar.AI.supports?/2` - Check specific capability

- Comprehensive test suite

- Full documentation with examples



### Features



- Protocol-based design for maximum flexibility

- Runtime capability detection via protocol introspection

- Unified interface for multiple AI providers

- Automatic response normalization across providers

- Composite adapter with retry logic and automatic fallback

- Telemetry events for monitoring and debugging

- Mock adapter with call tracking for testing

- Type-safe API with comprehensive typespecs

- Heuristic fallback for basic functionality without AI



[0.1.0]: https://github.com/nshkrdotcom/altar_ai/releases/tag/v0.1.0