v1.0.11

What's Changed

• add record_to_memory before generate return reply_message by @Osier-Yi in #1039
• fix(werewolf game): Broadcast game result when game ends at night phase by @qbc2016 in #1065
• fix(openai formatter): Correctly extract image extension from URLs with query parameters by @qbc2016 in #1066
• feat(agent): Add A2A protocol support for inter-agent communication by @Sunrisea in #1027
• raise exception on MCP disconnection by @MengsD in #1024
• fix(gemini model): flatten json schema in tools for gemini chat model by @qbc2016 in #1050
• bugfix: parse anthropic response content block type logic error. by @FFengIll in #1070
• chore(version): update version to 1.0.11 by @DavdGao in #1074

New Contributors

• @Sunrisea made their first contribution in #1027
• @MengsD made their first contribution in #1024
• @FFengIll made their first contribution in #1070

Full Changelog: v1.0.10...v1.0.11

v1.0.10

What's Changed

• fix(version): update version to 1.0.9 by @DavdGao in #1009
• fix(docs): fix bug in the tts tutorial by @DavdGao in #1010
• fix(opentelemetry): fix versions for opentelemetry by @qbc2016 in #1011
• fix(openaichatmodel): fix a bug when choice.delta is None by @qbc2016 in #1015
• refactor(planner): refactor the meta planner example with the PlanNotebook class by @DavdGao in #901
• feat(tracing): support list input in _get_agent_messages and add tests by @zhijianma in #1013
• feat(deployment): add example for deploying a routing agent as a chat endpoint by @DavdGao in #911
• feat(tool): add original name in RegisteredToolFunction by @qbc2016 in #1028
• perf(evaluation): save task metadata and token usage in the evaluation dir by @DavdGao in #961
• fix mem0ai version <=0.1.116 by @Osier-Yi in #1042
• fix(react agent): filter None values when joining message text content by @qbc2016 in #1041
• docs(studio): update to latest studio Gifs by @zhijianma in #1034
• fix(formatter unittest): fix the issue for tool result id mismatch by @qbc2016 in #1043
• fix(react agent): hotfix parameter issues for handle_interrupt by @qbc2016 in #1049
• chore(version): update version to 1.0.10 by @DavdGao in #1051

Full Changelog: v1.0.9...v1.0.10

v1.0.9

What's Changed

• fix(docs): fix the typo of the incorrect port number in the tutorial by @VXenomac in #954
• feat(agent-skill): support Anthropic agent skills in ReActAgent and Toolkit class by @DavdGao in #952
• fix(model & doc): unify client_args and client_kwargs arguments in the chat model classes into client_kwargs by @qbc2016 in #957
• fix: Update plan change hook type to support awaitables by @lixiaofei123 in #940
• fix(toolkit): fix meta tool failing to deactivate tool groups by @DavdGao in #962
• feat(dashscopeformatter): Support video block for dashscope by @qbc2016 in #964
• fix(config):fix created_at timestamp format by @zhijianma in #976
• (feature) Reme short-term mem by @jinliyl in #967
• fix(stream): Add raising error in stream_printing_messages by @qbc2016 in #981
• add memory with compression by @Osier-Yi in #951
• fix(prompt): fix the prompt in werewolf game by @qbc2016 in #993
• fix(model): unfiorm kwargs in ollama and dashscope model by @qbc2016 in #986
• feat: Add py.typed to enable static type checking support by @CharlieZhao95 in #988
• fix(react agent): fix the logic check between the meta tool and Plannote by @qbc2016 in #995
• feat(agent): support to disable console output from environment variable by @DavdGao in #1001
• [hotfix] fix the response error in deep research agent by @Luohh5 in #1000
• fix(dashscopechatmodel): warning once for tool choice == required by @qbc2016 in #950
• feat(tracing): support OTel GenAI semantic conventions by @zhijianma in #892
• fix(repair json): hotfix for _json_loads_with_repair by @qbc2016 in #1002
• fix(DeepSeekChatFormatter): Add reasoning_content in messages for deepseek-reasoner by @qbc2016 in #998
• docs(tracing): add Alibaba Cloud CloudMonitor integration by @sipercai in #978
• feat(tts): implement tts by @qbc2016 in #965

New Contributors

• @VXenomac made their first contribution in #954
• @lixiaofei123 made their first contribution in #940
• @CharlieZhao95 made their first contribution in #988
• @sipercai made their first contribution in #978

Full Changelog: v1.0.8...v1.0.9

v1.0.8

Important Change

• The generate_response function in ReActAgent class is refactored in v1.0.8 (#910)
• The config is refactored with ContextVar for thread-safe (#933)

What's Changed

• docs(README): update README.md and README_zh.md with update news by @DavdGao in #899
• fix(pipeline): use Sequence for covariant type substitution in MsgHub by @2725244134 in #909
• refactor(agent): refactor the structured output logic within the ReActAgent class by @qbc2016 in #910
• fix(docs): fix the typos in the docstring by @qbc2016 in #919
• fix(agent): set tool choice to "required" when structured output by @qbc2016 in #928
• (feature)update reme_ai from 312 to 310 by @jinliyl in #924
• fix(docs): fix the bug in building tutorial by @DavdGao in #937
• feat(memory): add limit parameter in base LongTermMemroyBase class by @Osier-Yi in #913
• fix(tool): execute_python_code utf8 encode for Chinese codes&comments issue#915 by @shaoeric in #916
• fix(reme): fix the error when creating ReMePersonalLongTermMemory using OpenAITextEmebdding by @tianqihou in #929
• feat(toolkit): support different strategies for duplicated tools by @DavdGao in #939
• refactor(config): support thread and async safe configuration by using ContextVar by @zhijianma in #933
• fix(toolkit): merge $defs when extending tool function schema with nested models by @qbc2016 in #941
• fix the case when mem0LongtermMemory contains graphstore by @Osier-Yi in #942
• fix(formatter): implement image extraction from tool results by @qbc2016 in #938

New Contributors

• @2725244134 made their first contribution in #909
• @shaoeric made their first contribution in #916
• @tianqihou made their first contribution in #929

Full Changelog: v1.0.7...v1.0.8

v1.0.7

What's Changed

• fix(stream_printing_messages): fix the bug in stream_printing_messages pipeline by @DavdGao in #866
• Fix planner state by @Young198902 in #863
• Change ollama embedding api by @qbc2016 in #874
• fix(model): fix the bug that Qwen-Omni requires different input_audio format than OpenAI models by @qbc2016 in #882
• feat(memory): add reme long-term memory by @jinliyl in #865
• feat(rag): implement word reader in the RAG module by @qbc2016 in #852
• fix(agent): raise the CancelError when interruption happens during the tool execution by @DavdGao in #888
• fix(agent): separate the logic of the arguments enable_meta_tool and plan_notebook by @ceshihao in #880
• Add tune module to support training agent workflow with RL by @pan-x-c in #816
• Refine building tutorial by @qbc2016 in #879
• docs(contributing): add CONTRIBUTING.md; adjust the example directories by @DavdGao in #890
• fix(example): handle the wrong input type of the tool_names parameter in the meta planner example by @qbc2016 in #870
• ci(copilot): setup GitHub copilot instructions. by @DavdGao in #897
• feat(toolkit): support async func for the postprocess_func argument in toolkit by @qbc2016 in #885
• fix(plan, rag) fix the bug in the word reader and plan notebook class. by @DavdGao in #895

New Contributors

• @jinliyl made their first contribution in #865
• @ceshihao made their first contribution in #880

Full Changelog: v1.0.6...v1.0.7

v1.0.6

What's Changed

• [HotFix] Fix missing await in browser agent's format by @qbc2016 in #836
• Improve the werewolves game example in AgentScope by @DavdGao in #840
• Refine tutorial by @qbc2016 in #843
• Added a demo example showing how AgentScope can connect to Alibaba Cl… by @CodeSpaceiiii in #805
• Add uv package manager support by @qbc2016 in #847
• [HotFix] Fix no api key bug in DashScopeMultiModalEmbedding by @qbc2016 in #846
• Fix the bug use stream_printing_messages will print the generate_response tool_use message by @Young198902 in #831
• feat: Implement audio playback from URL by @SyedaAnshrahGillani in #822
• Add support for the mcp.types.EmbeddedResource with text resource by @ZiTao-Li in #839
• Add support for milvus lite vector database in the RAG module of AgentScope by @AIGCool in #825
• [Hotfix] Fix the key "dimension" in dashscope text embedding api by @qbc2016 in #856
• Hotfix mem0 by @Osier-Yi in #854
• fix(plan): fix the bug in plan notebook and support to convert string index into int automatically by @DavdGao in #858

New Contributors

• @CodeSpaceiiii made their first contribution in #805
• @Young198902 made their first contribution in #831
• @SyedaAnshrahGillani made their first contribution in #822
• @AIGCool made their first contribution in #825

Full Changelog: v1.0.5...v1.0.6

v1.0.5

What's Changed

• Update README with tool agent examples by @cuiyuebing in #813
• Support to obtain the printing messages from an agent instance as an async generator by @DavdGao in #814
• Refine tutorial by @qbc2016 in #804
• [HOTFIX]: Add default value None for memory_type in _mem0_record by @zhijianma in #820
• Fix the bug in plan notebook serialization by @DavdGao in #826
• Update version number to 0.1.5 by @DavdGao in #827

New Contributors

• @cuiyuebing made their first contribution in #813

Full Changelog: v1.0.4...v1.0.5

v1.0.4

What's Changed

• Support voice agent in AgentScope by supporting Qwen-Omni and GPT-audio models
• Create RAG module in AgentScope

Commits

• Fix parallel tool execution in tutorial by @qbc2016 in #789
• [HOTFIX] Fix the bug in ReActAgent class by @DavdGao in #788
• [Hotfix] Fix ThinkingBlock import issue by @qbc2016 in #794
• fix(model): 修复DashScopeChatModel enable_thinking 不起作用 by @allenli178 in #796
• Support omni series model by @qbc2016 in #759
• Update version to 1.0.4 by @DavdGao in #798
• Add RAG module into AgentScope by @DavdGao in #800

New Contributors

• @allenli178 made their first contribution in #796

Full Changelog: v1.0.3...v1.0.4

v1.0.3

Important Change

• A new plan module is added
• The session_id argument in JSONSession class is moved into the load and save function to support multiple sessions.

What's Changed

• Add concurrent example by @qbc2016 in #693
• Add example for SQLite-based session management by @DavdGao in #740
• [HOTFIX] Fix the studio-related error in the tutorial by @DavdGao in #741
• [HOTFIX] Fix the bug and simplify the MCP example by @DavdGao in #754
• Fix tutorial format by @qbc2016 in #751
• fix: correct handling of non-existent file in insert_text_file by @idontwanttosayaword in #735
• fix bug of self._agent_control by @Osier-Yi in #760
• Separate the "plan" from the meta planner example and make it an independent module by @DavdGao in #722
• Make the metric call function async by @ZiTao-Li in #758
• Fix Gemini formatter local file mime_type from image/* to proper medi… by @idontwanttosayaword in #736
• [HOTFIX] Fix the bug in the plan module by @DavdGao in #771
• Remove the argument session_id from the constructor to support different sessions by @DavdGao in #767
• Fix: Missing Vllm token usage info by @Howe829 in #769
• [HOTFIX] modify the ray import in evaluator by @ZiTao-Li in #781
• Remove some duplicated requirements by @Howe829 in #777
• [HOTFIX] Fix the bug of missing metadata field when finish function fail to execute in the toolkit by @DavdGao in #783

New Contributors

• @Howe829 made their first contribution in #769

Full Changelog: v1.0.2...v1.0.3

v1.0.2

What's Changed

• [hotfix] fix error in deep research agent by @Luohh5 in #704
• Refine tutorial for multi mcp stateful clients by @qbc2016 in #706
• Add news for agentscope runtime by @rayrayraykk in #712
• Add tutorial for fanout pipeline by @qbc2016 in #700
• feat: improve logging setup by validating level by @Mirza-Samad-Ahmed-Baig in #715
• Fix typing for content blocks and Gemini formatter annotations by @idontwanttosayaword in #719
• Fix logo size in tutorial by @qbc2016 in #713
• Add the missing dependencies; update version; update mcp version by @DavdGao in #734

New Contributors

• @Mirza-Samad-Ahmed-Baig made their first contribution in #715
• @idontwanttosayaword made their first contribution in #719

Full Changelog: v1.0.1...v1.0.2