mirror of https://github.com/dapr/dapr-agents.git
7fac47b6d4
* Refactored Workflow & Task Decorators for Easier Standalone Registration * Refactored WorkflowApp: Improved Task & Workflow Registration, Enhanced State Management * Refactored WorkflowTask: Improved LLM Execution, Validation, and Chat History Support * Refactored API Server & Workflow Services: Unified FastAPI & Dapr Integration * Refactored Pub/Sub and Message Routing: Unified Dapr Integration & CloudEvent Parsing * Introduced AgenticWorkflowService: Unified Agent Workflows, State Management & Messaging * Enhanced Orchestrators with Dapr Workflow Patterns & Intelligent Message Routing * Updated Text Printer for Agents to show their name. * Introduced Assistant Agent with Tool-Driven LLM Workflow, Dapr Messaging, and Adaptive Execution * Consolidated Actor Wrappers into Unified Agent Actor Service and updated agent response logic * Removed workflows folder from Actors directory. Workflow App and Service folders taking over. * Introduced Execution Framework with Docker and Local Code Executors for Agents. * Refactored Dapr gRPC Configuration: Unified Naming and Improved Initialization Logic * Updated VectorStoreBase to Use Explicit EmbedderBase Type Annotation * Enhanced Structured Response Handling: JSON Schema & Iterable Support * Ran notebooks with latest version. * Updated agentic workflow docs * removed floki from huggingface chat client * Updated init files for latest changes * updated dapr pip package to 1.5.0 to use latest version * removed floki reference from code and docs. Tags in README still point to floki * Added a dummy notebook to test ReAct Agents |
||
|---|---|---|
| .. | ||
| neo4j_basic_examples.ipynb | ||