1. Add the "Skills" module;
2. The loading of the model square has been modified to be controlled through environment variables;
3. Dynamic scheduling of the skill binding tool;
4. Agent Integration Skills
* [changes]Add 'id' as the secondary sorting key, and 'scene_id' now returns a UUID object
* [fix]Fix the "end_user" return to be sorted by update time.
* [fix]Set the default values of the memory configuration model based on the spatial model.
* [fix]Remove the entity extraction check combination model, read the configuration list, and add the return of scene_id
* [fix]Fix the "end_user" return to be sorted by update time.
* [fix]
* [fix]Fix the issue of inconsistent language in explicit and episodic memory.
* [fix]Fix the issue of inconsistent language in explicit and episodic memory.
* [add]Add scene_id
* [fix]Based on the AI review to fix the code
* [modify] migration script
* [add] migration script
* fix(web): change form message
* fix(web): the memoryContent field is compatible with numbers and strings
* feat(web): code node hidden
* fix(model):
1. create a basic model to check if the name and provider are duplicated.
2. The result shows error models because the provider created API Keys for all matching models.
---------
Co-authored-by: Mark <zhuwenhui5566@163.com>
Co-authored-by: zhaoying <yzhao96@best-inc.com>
Co-authored-by: yingzhao <zhaoyingyz@126.com>
Co-authored-by: Timebomb2018 <18868801967@163.com>
1. create a basic model to check if the name and provider are duplicated.
2. The result shows error models because the provider created API Keys for all matching models.
* [fix]Fix the interface for statistics of recent activities and applications
* [changes]Modify the code based on the AI review
1.Use the boolean auxiliary methods provided by SQLAlchemy instead of using == True in the is_active filter.
2.The calculation of the "PROJECT_ROOT" has now been hardcoded with five levels of nested os.path.dirname calls.
* [fix]Fix the interface for statistics of recent activities and applications
* [changes]Modify the code based on the AI review
1.Use the boolean auxiliary methods provided by SQLAlchemy instead of using == True in the is_active filter.
2.The calculation of the "PROJECT_ROOT" has now been hardcoded with five levels of nested os.path.dirname calls.
* [changes]《Modify the interface》
1.Remove the "/search/entity_graph" interface
2.Reconstruct the "/updated_end_user/profile" interface
3.Remove the "Update Username" interface
4.Fix the batch query of user association memory configuration
* [changes]《Modify the interface》
1.Remove the "/search/entity_graph" interface
2.Reconstruct the "/updated_end_user/profile" interface
3.Remove the "Update Username" interface
4.Fix the batch query of user association memory configuration
* [fix]Fix the error response type
* refactor(celery): optimize task routing and worker configuration
- Simplify Celery queue configuration with single default 'io_tasks' queue
- Implement task routing strategy separating IO-bound and CPU-bound tasks
- Add Flower monitoring support with task event tracking enabled
- Add summary node search optimization to only retrieve summary nodes
- Clean up unused imports and reorganize import statements for consistency
- Update docker-compose configuration to support multi-queue worker setup
* chore(celery): simplify flower configuration and add gevent dependency
* chore(dependencies): add gevent dependency to requirements
- Add gevent==24.11.1 to api/requirements.txt
- Gevent is required for async worker support in Celery
- Complements existing flower and celery configuration
* refactor(celery): simplify async event loop handling and reorganize task queues
- Replace complex nest_asyncio and manual event loop management with asyncio.run() in read_message_task, write_message_task, regenerate_memory_cache, and workspace_reflection_task
- Rename task queues from io_tasks/cpu_tasks to memory_tasks/document_tasks for better semantic clarity
- Update task routing configuration to reflect new queue names for memory agent tasks and document processing tasks
- Remove redundant exception handling comments and simplify error handling logic
- Update README with improved community support section including GitHub Issues, Pull Requests, Discussions, and WeChat community links
- Simplifies event loop management by leveraging asyncio.run() which handles loop creation and cleanup automatically, reducing code complexity and potential race conditions
* [feature]A set of information for role recognition writing
* [feature]A set of information for role recognition writing
* [fix]Fix the code after rebasing.
* [feature]A set of information for role recognition writing
* [fix]Fix the code after rebasing.
* [fix]Based on the AI review to fix the code
* [changes]Disable the function of batch writing multiple groups of conversations in a cumulative manner
* [fix]Addressing vulnerability risks
* [feature]Emotional memory cache
* [feature]Implicit memory cache
* [changes]Modify the expiration time of implicit memory to 24 hours.
* [feature]Emotional memory cache
* [feature]Implicit memory cache
* [changes]Modify the expiration time of implicit memory to 24 hours.
* [changes]Modify the code based on the AI review
* [feature]Emotional memory cache
* [feature]Implicit memory cache
* [changes]Modify the expiration time of implicit memory to 24 hours.
* [feature]Implicit memory cache
* [changes]Modify the code based on the AI review
* [fix]Correct the display sequence of memory increments
* [fix]Correct the display sequence of memory increments
* [changes]Modify the code based on the AI review
* [feature]Generate emotions, implicit cache
* [feature]Generate emotions, implicit cache
* [changes]Improve the code based on AI review
* [changes]Improve the code based on AI review
* [changes]Improve the code
* [feature]Generate emotions, implicit cache
* [changes]Improve the code based on AI review
* [changes]Improve the code
* [fix]Fix the return of the "content" attribute
* [changes]Improve the code based on AI review
* Apply suggestion from @sourcery-ai[bot]
Co-authored-by: sourcery-ai[bot] <58596630+sourcery-ai[bot]@users.noreply.github.com>
* [fix]Fix the return of the "content" attribute
* [changes]Improve the code based on AI review
* Apply suggestion from @sourcery-ai[bot]
Co-authored-by: sourcery-ai[bot] <58596630+sourcery-ai[bot]@users.noreply.github.com>
* [changes]Improve the code based on AI review
---------
Co-authored-by: sourcery-ai[bot] <58596630+sourcery-ai[bot]@users.noreply.github.com>
* refactor(conversation): separate service and repository layers for conversation module
- Split ConversationService and repository/UnitOfWork layers
- Service layer now only handles business logic and orchestration
- Repository layer handles all direct database operations
- UnitOfWork encapsulates transactional operations for messages
- Ensured all public methods have clear English docstrings with arguments, return values, and exceptions
* feat(memory): implement work memory endpoints and services
- Added API routes for conversation count, conversation list, messages, and detail.
- Integrated ConversationService for database queries and LLM-based summary generation.
* feat(memory): implement work memory endpoints and services
- Added API routes for conversation count, conversation list, messages, and detail.
- Integrated ConversationService for database queries and LLM-based summary generation.
* feat(workflow): fix issues causing workflow failures
if-else None value error
knowledge empty list rerank
end node output none node value
assigner input none value
* feat(memory): convert memory file creation time to timestamp and include title and first-line fields in file type
* fix(memory): fix serialization output and default value issues
* fix(workflow): fix issue with hybrid search logic in knowledge retrieval node
* [changes]Request to remove 'config_id' has been received.
* [add]Add the access history record table
* [changes]Request to remove 'config_id' has been received.
* [add]Add the access history record table
* [add]Obtain the record of the forgetting trend
* [changes]Based on the AI's suggestion, make the necessary modifications.
* perf(workflow): pass JSON data to HTTP node as a string
* perf(prompt_opt): simplify log output
* feat(memory): add perceptual memory page API and related database schema
* perf(log): clean up API exception log output
* perf(memory): simplify perceptual memory timeline response by removing metadata
