Name | Modified | Size | Downloads / Week |
---|---|---|---|
Parent folder | |||
ai-gateway-aarch64 | 2025-07-08 | 35.6 MB | |
ai-gateway-x86_64 | 2025-07-08 | 40.4 MB | |
langdb_udf-aarch64 | 2025-07-08 | 7.8 MB | |
langdb_udf-x86_64 | 2025-07-08 | 8.8 MB | |
README.md | 2025-07-08 | 4.0 kB | |
Release 0.2.3 source code.tar.gz | 2025-07-08 | 43.9 MB | |
Release 0.2.3 source code.zip | 2025-07-08 | 44.0 MB | |
Totals: 7 Items | 180.4 MB | 0 |
Changelog
All notable changes to this project will be documented in this file. See standard-version for commit guidelines.
0.2.3 (2025-07-08)
Features
- add custom event for model events (3d34406)
- add description and keywords fields to thread (566512c)
- Add key generation for transport type (7d15fc9)
- Add options struct for prompt caching (7cdf8d3)
- add run lifecycle events and fix model usage tracking (5b86f23)
- Add variables field to chat completions (#73) (ae94f87)
- add version support for virtual model retrieval via model@version syntax (8c26bb3)
- Basic responses support (#92) (67efe60)
- Enhanced support for MCP servers (#72) (46137d1)
- Handle max retries in request (ccb5f0a)
- implement tenant-aware OpenTelemetry trace (034d15d)
- Support azure url parsing and usage in client (363dd75)
- Support http streamable transport (#86) (ed7ca6e)
- Support project traces channels (e9e6928)
Bug Fixes
- Empty required parameters list (5e8c0cd)
- Fix duplicated tools labels in gemini tools spans (7c01c7a)
- Fix operation name for model spans (1f53ef2)
- Fix required default value (3d5bd55)
- Fix retries handle in llm calls (64fd662)
- Fix retries logic (94bccf1)
- Fix tracing for cached responses (e6f0171)
- handle cache response errors gracefully in gateway service (8900fab)
- Handle thought signature in gemini response (3f0116d)
- Properly handle model calls traces in gemini (707d599)