@revenium/openai 1.0.11 → 1.0.13

package/.env.example

@@ -0,0 +1,20 @@
+# Revenium OpenAI Middleware Configuration
+# Copy this file to .env and fill in your actual values
+
+# Required: Your Revenium API key (starts with hak_)
+REVENIUM_METERING_API_KEY=hak_your_revenium_api_key_here
+
+# Optional: Revenium API base URL (defaults to https://api.revenium.ai)
+#REVENIUM_METERING_BASE_URL=https://api.revenium.ai
+
+# Required: Your OpenAI API key (starts with sk-)
+OPENAI_API_KEY=sk_your_openai_api_key_here
+
+# Optional: Azure OpenAI configuration (uncomment if using Azure)
+#AZURE_OPENAI_ENDPOINT=https://your-resource-name.openai.azure.com/
+#AZURE_OPENAI_API_KEY=your-azure-openai-api-key-here
+#AZURE_OPENAI_DEPLOYMENT=your-deployment-name-here
+#AZURE_OPENAI_API_VERSION=2024-12-01-preview
+
+# Optional: Enable debug logging
+#REVENIUM_DEBUG=true

package/CHANGELOG.md

@@ -5,73 +5,73 @@ 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).
 
-## [1.0.11] - 2025-10-21
+## [1.0.13] - 2025-11-06
 
 ### Added
-- Examples now included in npm package for offline access
-- New `examples/README.md` with comprehensive step-by-step guide
-- Step 4: Protect Your API Keys - Security best practices in main README
-- TypeScript and JavaScript examples for getting started
+- configure() function for manual configuration (simpler alias for initializeRevenium)
+- Support for COST_LIMIT and COMPLETION_LIMIT stop reasons
+- Support for all 7 operationType enum values (CHAT, GENERATE, EMBED, CLASSIFY, SUMMARIZE, TRANSLATE, OTHER)
 
 ### Changed
-- Removed internal QA environment references from documentation
-- Documentation standardized for public repository
-- Improved README structure with enhanced security guidance
-
-### Security
-- Added `.gitignore` best practices to protect API keys
-- Enhanced environment variable security documentation
+- **BREAKING:** responseQualityScore now uses 0.0-1.0 scale (was 0-100) per updated API specification
+- API endpoint updated from api.revenium.io to api.revenium.ai (new production domain)
+- Improved API compliance with proper enum values and field casing
+- Enhanced token counting accuracy for cache operations
+- Email addresses now masked in debug logs for PII protection
+- All examples updated to use 0.0-1.0 scale for responseQualityScore
+
+### Fixed
+- Improved documentation accuracy and removed deprecated configuration options
+- config.debug property now properly enables debug logging
+- Manual configuration example now includes required patchOpenAIInstance call
+- Updated OpenAI version requirement to match peer dependency (v5.0.0+)
+- Enhanced metadata validation for responseQualityScore scale
+- Improved provider detection fallback handling
+- Fixed cache token reporting for embeddings API
+- reasoningTokenCount now correctly optional based on provider support
+
+## [1.0.12] - 2025-10-30
 
-## [1.0.10] - 2025-10-17
+### Changed
+- Improved documentation structure and Getting Started tutorial
+- Enhanced package distribution to include TypeScript source files
 
-### Added
-- Support for OpenAI Responses API (SDK 5.8+)
-- New example files for Responses API with Azure and standard OpenAI
-- Enhanced TypeScript type definitions for Responses API
+### Fixed
+- Repository configuration updates
 
-### Changed
-- README documentation significantly expanded with step-by-step guides
-- Improved examples structure and organization
-- Package maintenance and stability improvements
+## [1.0.11] - 2025-10-21
 
-## [1.0.9] - 2025-10-16
+### Added
+- Examples now included in npm package
+- Comprehensive examples documentation
 
 ### Changed
-- Internal improvements and bug fixes
-- Enhanced error handling and logging
+- Updated API endpoint configuration
+- Improved documentation structure
 
-## [1.0.8] - 2025-10-15
+## [1.0.10] - 2025-10-17
 
 ### Added
-- Initial release of @revenium/openai scoped package
-- Migration from revenium-middleware-openai-node to @revenium/openai
-- Transparent OpenAI API usage tracking middleware
-- Support for streaming responses
-- Support for function/tool calling
-- Automatic metadata collection and forwarding
-- Azure OpenAI support with automatic detection
-- Full TypeScript support with native type integration
-- Zero configuration setup from environment variables
+- Support for OpenAI Responses API
 
 ### Changed
-- Package name migrated to scoped format (@revenium/openai)
-
----
-
-## Package Migration Notice
+- Enhanced documentation and examples
 
-This package was previously published as `revenium-middleware-openai-node` up to v1.1.0.
+## [1.0.9] - 2025-10-16
 
-Starting with v1.0.8, the package has been migrated to the scoped format `@revenium/openai`.
+### Changed
+- Bug fixes and stability improvements
 
-The old package has been deprecated with a notice directing users to the new scoped package.
+## [1.0.8] - 2025-10-15
 
-To upgrade:
-```bash
-npm uninstall revenium-middleware-openai-node
-npm install @revenium/openai
-```
+### Added
+- Initial release with OpenAI usage tracking
+- Support for streaming responses and function calling
+- Azure OpenAI support
+- TypeScript support with native type integration
 
+[1.0.13]: https://github.com/revenium/revenium-middleware-openai-node/releases/tag/v1.0.13
+[1.0.12]: https://github.com/revenium/revenium-middleware-openai-node/releases/tag/v1.0.12
 [1.0.11]: https://github.com/revenium/revenium-middleware-openai-node/releases/tag/v1.0.11
 [1.0.10]: https://github.com/revenium/revenium-middleware-openai-node/releases/tag/v1.0.10
 [1.0.9]: https://github.com/revenium/revenium-middleware-openai-node/releases/tag/v1.0.9