CHANGELOG.md

Changelog

[0.1.4] - 2025-01-25

✨ New Features

• Added Gemini AI provider support with token counting
• Added Getting Started guide with installation instructions
• Added audit logs documentation with example log entries

💄 UI Changes

• Updated logo across documentation and README
• Enhanced sequence diagram with emojis for better visualization
• Improved documentation navigation structure

🔧 Technical Changes

• Added support for API keys in proxy URLs
• Implemented configurable authorization header toggle
• Enhanced URL handling with proper query parameter support

📝 Documentation

• Updated main landing page with new features
• Added new sections in documentation:
  • Getting Started guide
  • Features overview
  • Audit logging details

🎨 Assets

• Added new logo variations

🔄 Updates & Modifications

• Reorganized documentation structure
• Updated configuration example with Gemini provider
• Added extra newlines in configuration file for better readability

[0.1.3] - 2025-01-24

🛠️ Code Improvements & Bug Fixes

• Audit Log Rotation 🔄
  • Updated the audit log configuration (log4rs.yml) to support rotating logs based on size (50MB limit) and date.
  • Added a fixed-window roller to maintain up to 5 archived log files.
  • Ensured logs are stored persistently with configurable retention periods.
• Error Handling in Audit Logs 🚨
  • Added error logging for responses with HTTP status codes >= 400 in gateway.rs.
  • Improved request logging by removing newline characters for cleaner audit logs.
• Configuration File Handling ⚙️
  • Fixed an issue in main.rs where the program would not exit properly when no configuration file was provided.
  • Removed redundant env_logger::init() call and added a proper error message.

---

📚 Documentation Updates

• Features Documentation 📝
  • Added a new section in the documentation (mkdocs.yml) to detail the Features of the gateway, including audit logs and storage.
• README.md Updates 📖
  • Added descriptions for Audit Logs and Audit Storage features in the README.

---

🚀 New Features

• Audit Log Enhancements 📋
  • Introduced detailed logging of API requests and responses in gateway.rs.
  • Added persistent storage for all gateway exchanges with configurable retention periods.

---

🧹 Code Cleanup

• Removed redundant code and improved readability in gateway.rs and main.rs.

[0.1.2] - 2025-01-23

🛠️ Fixes

• JSON Handling:
  • Fixed JSON escaping by replacing manual handling with jq for proper character handling. (024c6f0)
  • Sanitized ACTIVITY_CONTENT for JSON by escaping special characters. (a6dc37d)
• CORS Improvements:
  • Added CORS preflight request handling in the gateway. (f1d1d94)
  • Removed no-cors mode and enhanced CORS headers for proper response handling. (c210d41)
• Code Cleanup:
  • Simplified parser assignment and removed redundant code in chat.html. (1949cd2)
  • Simplified configuration handling and response filtering in the gateway. (4d3dd85)
• Miscellaneous Fixes:
  • Various fixes applied. (f81fa3e)

✨ Features

• Automation:
  • Added activity4changelog.sh script to automate changelog generation. (11283a4)
• Chat Service Enhancements:
  • Added chat service with conversation management endpoints. (a2409c0)
  • Delegated chat API handling to the admin module. (1c74add)
  • Added default host and port configurations for chat and echo services. (bfb6610, 815cf9b)
• UI/UX Improvements:
  • Added a save button to the navbar for explicit conversation saving. (fdc1b63)
  • Displayed individual conversation items in the sidebar with previews. (d5691c5)
  • Made the navbar fixed and full-width with proper content spacing. (ed9cc13)
  • Added Enter key message sending with Shift+Enter for new lines. (569ed9c)
  • Added model selection dropdown to the chat interface. (7d49b8b)
• API Enhancements:
  • Updated model selection to handle new API response format. (7bab77f)
  • Added model-specific message parsers in parsers.js. (3ae29f0)
  • Added Authorization header with bearer token to API requests. (3477fb4)
  • Dynamically populated model-select from gateway endpoint. (e3b5028)
• Configuration:
  • Handled root path requests by returning configuration in JSON format. (e80b498)
  • Added early return for root and login paths with YAML config. (e1fdf62)

📚 Documentation

• Added chat interface documentation and navigation. (11283a4)

🔄 Refactors

• Code Simplification:
  • Used composition to delegate common functionality to HttpAdminApp. (a2408a8)
  • Simplified error messages in chat.html. (113168e)
  • Simplified request handling and moved root path logic to response filter. (0eece05)
• UI Refactoring:
  • Moved inline styles to external CSS and adjusted navbar layout. (1456084)
  • Replaced navbar links with server URL input in chat.html. (41c5eba)

🧹 Chores

• Updated proxy_pass URL in conf.yml to use port 6193. (159f612)
• Added service declaration route for the / URI to display all available models. (d1830be)

🎨 Styling

• UI Adjustments:
  • Adjusted model-select margin for left alignment. (a997bbb)
  • Aligned model-select to the left in input-form and adjusted spacing. (a6fcbcd)
  • Adjusted widths for #model-select and #send-button elements. (ebdaf30)
  • Adjusted form element widths and spacing in input-form. (701f135)
  • Adjusted navbar image margin and chat container alignment. (29a7263)
• Color & Layout:
  • Updated chat interface colors for better readability. (b23df74)
  • Commented out height calculation in CSS. (689ee22)

✨ New Features & Enhancements

• OpenAI API Integration:
  • Added support for OpenAI API integration.
  • Example endpoint: /api.openai.com/v1/chat/completions.
  • Parser added for OpenAI token usage tracking.
• Gzip Encoding Support:
  • Added gzip decoding for upstream responses.
  • Ensures efficient data transfer and processing.
• Admin Service Configuration:
  • Added admin host/port configuration to server config.
  • Improved flexibility for deployment environments.
• CI/CD Pipeline:
  • Set up GitHub Actions for continuous integration.
  • Added workflows for building, testing, and releasing Rust binaries.
  • Supports multiple platforms: Linux (x86_64, arm64) and macOS.

🛠️ Bug Fixes & Improvements

• Token Validation:
  • Improved token validation in the admin interface.
  • Tokens shorter than 32 characters are now rejected.
• Response Header Handling:
  • Fixed response header handling in the gateway.
  • Removed hardcoded headers and improved content encoding management.
• PII Protection:
  • Enhanced PII (Personally Identifiable Information) detection.
  • Added tests for long text inputs with PII.
• Echo Service:
  • Renamed request counter for clarity (REQ_COUNTER to ECHO_REQ_COUNTER).

🧪 Tests & Quality Assurance

• Admin Interface Tests:
  • Added comprehensive tests for token creation, deletion, and usage stats.
• Local Tests:
  • Added local tests for PII and long text handling.
• Remote Tests:
  • Expanded remote tests for Deepseek and OpenAI APIs.

📚 Configuration & Documentation

• Config Updates:
  • Updated conf.yml with new admin host/port settings.
• Changelog Updates:
  • Added details about new features, bug fixes, and tests.

🎉 Miscellaneous

• Renamed changelog.md to CHANGELOG.md for consistency.
• Added flate2 crate for gzip support.

---

[0.1.1] - 2025-01-23

Added

• ✅ Support to OpenAI API
• 🔧 Manage upstream content encoding with gzip
• 🛠️ Continuous integration with GitHub Actions

[0.1.0] - 2025-01-22

Added

• 🛠️ Complete Rust migration for improved performance and safety
• 🚪 New admin service with web UI for configuration and monitoring
• 📊 Prometheus metrics integration for real-time monitoring
• 🔑 Token management system with quotas and rate limiting
• 🤖 Support for multiple LLM providers (OpenAI, Claude, DeepSeek, Ollama)
• 🔒 PII detection and content filtering capabilities

Changed

• 📄 Complete documentation overhaul with new guides and API references
• 🔄 Refactored rate limiting system for better scalability
• 🛡️ Enhanced security with trusted header authentication

Removed

• 🗑️ Legacy Python implementation
• 🚫 Deprecated configuration options

---

This project follows Semantic Versioning