Conversation
…multi-modal agent example - Deleted outdated documentation for reporters and user simulation. - Added a new example demonstrating multi-modal agent workflows using various output types. - Refactored the react_sync example to utilize the Agent class for improved tool integration and context management.
- Added EvalPresets class for common evaluation configurations including response quality, tool usage, conversation flow, and comprehensive checks. - Implemented QuickEval class to streamline evaluation processes with methods for single checks, batch evaluations, and preset configurations. - Created GoogleEmbedding class for integrating Google embedding services with error handling and custom dimensionality support. - Developed QuickTest class for simplified testing of agents with built-in assertions for single-turn and multi-turn interactions, including tool usage. - Added example scripts demonstrating usage of QuickEval, EvalPresets, and embedding providers. - Introduced test result assertions for QuickTest to enhance testing capabilities.
- Created run_tests.py to facilitate running various test suites with options. - Added unit tests in test_react_sync.py for core functionalities of the react_sync module. - Implemented evaluation tests in test_react_sync_evaluation.py to validate agent performance and tool usage. - Introduced test_requirements.txt to manage dependencies for testing.
… app - Implemented eval_suite.py to provide a comprehensive evaluation framework with critical, tool, performance, and edge case tests. - Created evaluation.py to define metrics, evaluation cases, and the ReactEvaluator class for streamlined evaluation. - Introduced react_sync.py to set up the agent and tools, including a weather function and state management for tool usage. - Added detailed reporting and metrics aggregation to enhance evaluation insights.
…enAI as a dependency
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request introduces major improvements to agent communication and expands SDK support. The most significant change is the addition of a new agent-to-agent communication module, which provides a high-level API for messaging, broadcasting, and pub-sub patterns. There are also enhancements to LLM converter selection logic, cleanup of documentation and task planning files, and updates to the SDK feature checklist.
New Agent Communication Module:
agentflow/communication/a2a.py, implementing theA2ACommunicationManagerclass for agent-to-agent communication, including registration, messaging, broadcasting, pub-sub, heartbeat, and agent management APIs.agentflow/communication/__init__.pyto expose communication-related classes (A2ACommunicationManager,AgentRegistry,AgentRegistryEntry,MessageRouter) for external use.LLM Adapter Improvements:
agentflow/adapters/llm/model_response_converter.pyto support OpenAI, Google, and LiteLLM converters, and provide error handling for unsupported types.OpenAIConverterinagentflow/adapters/llm/__init__.py, making it available in the SDK.Documentation and Planning Cleanup:
TaskPlan.md, streamlining the task plan for future improvements.Checklist.mdto reflect current major SDK features and protocol support.Build and Test Scripts:
Makefileto simplify build and test scripts.