This release adds 3 notable features for engineering teams evaluating rollout.
✓ No known CVEs patched in this version
Topics
+2 more
Summary
AI summaryUpdates model, readme, and formatter across a mixed release.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Feature | Medium |
Adds reasoning_effort options for OpenAI-compatible providers in model component Adds reasoning_effort options for OpenAI-compatible providers in model component Source: llm_adapter@2026-05-25 Confidence: high |
— |
| Feature | Medium |
Adds support for OpenAI response API in model component Adds support for OpenAI response API in model component Source: llm_adapter@2026-05-25 Confidence: high |
— |
| Feature | Medium |
Adds thought signature for Gemini model and formatter in gemini component Adds thought signature for Gemini model and formatter in gemini component Source: llm_adapter@2026-05-25 Confidence: high |
— |
| Bugfix | Medium |
Fixes model handling of structured output failures on Dashscope‑compatible endpoints Fixes model handling of structured output failures on Dashscope‑compatible endpoints Source: llm_adapter@2026-05-25 Confidence: high |
— |
| Bugfix | Medium |
Fixes formatter emitting `text: null` in AnthropicChatFormatter tool_result Fixes formatter emitting `text: null` in AnthropicChatFormatter tool_result Source: llm_adapter@2026-05-25 Confidence: low |
— |
| Bugfix | Medium |
Fixes TTS bug, adds warning and retry handling Fixes TTS bug, adds warning and retry handling Source: llm_adapter@2026-05-25 Confidence: low |
— |
| Bugfix | Medium |
Fixes tracing to handle CancelledError when closing spans Fixes tracing to handle CancelledError when closing spans Source: llm_adapter@2026-05-25 Confidence: low |
— |
| Other | Medium |
Bumps version to 1.0.21 in version management Bumps version to 1.0.21 in version management Source: llm_adapter@2026-05-25 Confidence: low |
— |
Full changelog
What's Changed
- docs: update dingtalk qr codes and add hyperlink by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1620
- docs(readme): remove redundant links by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1636
- feat(model): extend reasoning_effort options for OpenAI-compatible providers by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1643
- fix(formatter): avoid emitting
text: nullin AnthropicChatFormatter tool_result by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1641 - feat(model): support openai response api by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1346
- fix(tts): bug fix and add warning and retry by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1014
- feat(gemini): add thought signature for gemini model and formatter by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1402
- fix(tracing): handle CancelledError when closing spans by @FallingSnowFlake in https://github.com/agentscope-ai/agentscope/pull/1619
- fix(model): handle structured output failures on dashscope-compatible endpoints by @CarlBye in https://github.com/agentscope-ai/agentscope/pull/1646
- chore(version): bumping version to 1.0.21 by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/1652
New Contributors
- @FallingSnowFlake made their first contribution in https://github.com/agentscope-ai/agentscope/pull/1619
- @CarlBye made their first contribution in https://github.com/agentscope-ai/agentscope/pull/1646
Full Changelog: https://github.com/agentscope-ai/agentscope/compare/v1.0.20...v1.0.21
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
Related context
Related tools
Beta — feedback welcome: [email protected]