Release 3.0.0

[battmanux]

This pull request introduces major documentation and feature updates for OpenHosta as part of the v3.0.0 release, with a focus on improved clarity, expanded installation instructions, and the addition of experimental features for future development. The most significant changes include a complete overhaul of the README.md and docs/installation.md to better guide users, updates to the changelog for model naming consistency, and the addition of a next directory containing prototypes for image handling and agent support.

Documentation Overhaul

• The README.md has been rewritten to provide clearer usage instructions, updated examples (including semantic tests and AI-powered function emulation), and simplified installation guidance, with improved explanations of local and remote model setup. [1] [2] [3]
• The docs/installation.md has been extensively revised to clarify installation via pip, detail additional dependencies for advanced features, explain API key and local model configuration, and provide troubleshooting tips. [1] [2]

Changelog and Model Naming Consistency

• The CHANGELOG.md entries have been updated to consistently use OpenAICompatibleModel instead of Model, reflecting changes in class naming throughout the project. This affects feature, fix, and documentation notes for recent versions. [1] [2] [3] [4] [5] [6]

Experimental Features for Next Release

• Added a next directory containing prototypes for future features:
  • ImgPipeline.py: Demonstrates image input handling and document classification using AI, with notes on model limitations and inspection utilities.
  • agents.py: Provides a sample smart home agent implementation, illustrating planned support for tool usage and context passing to LLMs.
  • README.md: Introduces the purpose of the next folder for upcoming features.

General Improvements

• Documentation and examples now consistently reference the latest version and include links to relevant test files and documentation for further exploration. [1] [2]

These changes collectively improve the onboarding experience, clarify advanced usage scenarios, and lay the groundwork for new capabilities in OpenHosta.

[battmanux]

This pull request introduces OpenHosta v3.0.0, which includes major documentation updates, improved installation instructions, and new example/test files for upcoming features. The main focus is on clarifying usage, expanding support for local and remote LLM models, and preparing for future agent and image processing functionalities. Most changes are centered around documentation (in README.md, CHANGELOG.md, and docs/installation.md) and the addition of new sample code in the next/ directory.

Documentation and Usage Improvements

• Updated README.md to reflect v3.0.0, with clearer descriptions, revised example usage, and improved instructions for setting up API keys and local models. Added semantic test and function emulation examples. [1] [2] [3]
• Overhauled docs/installation.md to clarify installation via pip, list additional dependencies for advanced features, and provide detailed instructions for setting up API keys and local models (Ollama). Improved troubleshooting and next steps guidance. [1] [2]

Changelog Updates

• Revised references in CHANGELOG.md from Model to OpenAICompatibleModel to accurately reflect class name changes and improve clarity for recent and past releases. [1] [2] [3] [4] [5] [6]

New Example and Test Files for Next Features

• Added next/ImgPipeline.py to demonstrate and test image handling capabilities with OpenHosta, including function emulation for image classification and markdown extraction. Contains notes on model compatibility and planned error handling for unsupported image models.
• Added next/agents.py to showcase a sample agent design for smart home control, illustrating how future versions will support tool invocation and local scope passing to LLMs.
• Created next/README.md as an index for upcoming features and experimental files.