Define and run multi-agent workflows in Python with first-class MCP support — chains, routers, parallel fan-out, evaluator-optimizer loops, and A2A interoperability.
- Access
- CLIAPIMCP
- Auth
- API Key
- Skills
- 3
Skills
Workflow Patterns
Composes chain, router, parallel, and orchestrator patterns as decorators over plain Python agent definitions.
MCP Integration
Attaches MCP servers to individual agents, including sampling, roots, and elicitation beyond basic tool calling.
Evaluator Loops
Runs generator and evaluator agents in a refinement loop until output reaches a defined quality rating.
Related Agents
Spring AI Alibaba
API · MCPBuilds agentic, workflow and multi-agent applications in Java and Spring Boot on a graph runtime for stateful agents…
mcp-agent
API · MCPBuilds agents from MCP servers using composable workflow patterns from the Building Effective Agents playbook, handling…
NVIDIA NeMo Agent Toolkit
API · CLI · MCPConnects, profiles, evaluates, and optimizes agent workflows across frameworks like LangChain, LlamaIndex, and CrewAI…
OpenAgents
API · CLIRuns a collaboration workspace where agents delegate to each other by mention, reachable through Python and TypeScript…