agent-sdk-go
by agenticenv
Go SDK for durable, protocol-ready AI agents with A2A and AG-UI support
Overview
Enables building AI agents in Go with support for durable workflows via Temporal or lightweight in-process execution. Implements A2A messaging protocol and MCP messaging, AG-UI components, streaming, RAG tool integration, sub-agents, and human-in-the-loop hooks to compose complex agent behaviors. Distinctive features include crash-resilient Temporal workflows and a Go-first SDK for production-grade concurrency and observability. This Go SDK aligns well with best practices like planning patterns to structure agent orchestration.
Key Benefits
Ideal For
Go teams building production-capable agents who need durable workflows (Temporal), A2A connectivity, and tooling integration for streaming and RAG. Leverage Tool Use Pattern to integrate external tools and coordinate actions across agents.
Real-World Examples
- Coordinate long-running agent workflows with Temporal to preserve state across crashes and restarts
- Build Go-native agents that communicate over A2A or MCP and record interaction logs for later reputation analysis
- Compose sub-agents and human-in-the-loop handoffs for reliable task delegation and observability
- Run lightweight in-process agents for testing or low-latency scenarios without external dependencies