Skip to content

Latest commit

 

History

History
42 lines (37 loc) · 2.42 KB

File metadata and controls

42 lines (37 loc) · 2.42 KB

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.

[0.0.1] - YYYY-MM-DD - MVP Release (PLANNED)

✨ Added

  • Initial MVP Release of the Allos Agent SDK!
  • Core Agent (allos.agent):
    • Implemented the core Agent class with a robust, multi-turn agentic loop.
    • Added AgentConfig for easy and type-safe agent configuration.
    • Implemented a human-in-the-loop permission system (ToolPermission) for secure tool execution.
    • Added full session management with agent.save_session() and Agent.load_session().
  • Provider Agnostic Layer (allos.providers):
    • Created a flexible provider abstraction (BaseProvider) and ProviderRegistry.
    • Added full support for the OpenAI Responses API.
    • Added full support for the Anthropic Messages API.
  • Tool System (allos.tools):
    • Created an extensible tool abstraction (BaseTool) and ToolRegistry with a @tool decorator for easy custom tool creation.
    • Added a comprehensive set of built-in tools:
      • read_file (with line range support)
      • write_file (with append mode)
      • edit_file (with uniqueness validation)
      • list_directory (with recursive support)
      • shell_exec (with timeouts and a security blocklist)
  • Command-Line Interface (allos.cli):
    • Developed a powerful and user-friendly allos CLI tool.
    • Supports direct task execution (e.g., allos "do something").
    • Includes a full-featured interactive mode (allos -i).
    • Provides commands for discovering tools and providers (--list-tools, --list-providers).
    • Supports all core agent features like provider switching, tool selection, session management, and auto-approval via flags.
  • Testing and Quality:
    • Achieved 100% unit test coverage for the core Python library.
    • Built comprehensive integration and E2E test suites.
    • Established a CI/CD pipeline for testing across multiple platforms (Linux, macOS, Windows) and Python versions (3.9-3.13).
  • Documentation and Examples:
    • Created a full documentation suite with guides, tutorials, and API references.
    • Provided a rich set of runnable examples for all major features.