swarms

Commit Graph

Author	SHA1	Message	Date
Pavan Kumar	cc56f433a8	fix(mcp): update architecture to initialize agents on the client side, not the server	3 months ago
Pavan Kumar	eef1a5384c	Restored to 'eae5d510b502852f51ba86a44012b42d892fd643' Replit-Restored-To: `eae5d510b5`	3 months ago
Pavan Kumar	c4eb44e67d	fix(agent): standardize agent initialization to match example pattern	3 months ago
Pavan Kumar	f69682f31b	chore: add detailed logging to MCP server for end-to-end data flow visibility - Instrumented `mock_math_server.py` to log incoming agent requests and tool responses - Ensured logs clarify the MCP protocol: agent → MCP server → tool → response → agent - Supports debugging and compliance with Model Context Protocol (MCP) flow	3 months ago
Pavan Kumar	40cd5248d8	fix: align math server tools and agent prompt with available capabilities - Updated system prompt in `mock_multi_agent.py` to reflect only supported tools - Added error handling in `mock_math_server.py` to reject unsupported operations - Ensured agent behavior stays within defined context and MCP protocol limits	3 months ago
Pavan Kumar	3899a22c97	fix: correct MCP agent response handling and model configuration - Set `gpt-4o-mini` explicitly as model to ensure consistent responses - Fixed agent configuration and response parsing in `mock_multi_agent.py` - Updated `mock_math_server.py` and `.replit` config for accurate debugging and logging	3 months ago
Pavan Kumar	a8ad884d27	feat: implement mock MCP server and multi-agent system with full integration flow - Created `mock_math_server.py` to simulate MCP server behavior - Developed `mock_multi_agent.py` to orchestrate agent–server interaction - Added `mock_integration_test.py` to validate end-to-end data flow - Ensured user input → server request → agent processing → user output pipeline	3 months ago

7 Commits (0654a297a2a388625aaaecd44cbf6e491da83381)