Build production Go agents with tRPC-Agent-Go
Model Go-native agent systems with graph workflows, tool calls, memory, MCP, A2A, evaluation, and OpenTelemetry observability.
npx skills add agentskillexchange/skills --skill build-production-go-agents-with-trpc-agent-go
Use tRPC-Agent-Go when an operator is building agent behavior inside a Go service and needs a production-shaped runtime instead of ad hoc LLM calls. The workflow is concrete: model the agent as a graph or runner, attach function tools or MCP tools, add session and memory state, expose integrations such as A2A or AG-UI when needed, then evaluate and trace runs with OpenTelemetry. Invoke this instead of using a product UI when the goal is to ship Go-native agent behavior as part of a backend system. The scope boundary is Go agent orchestration and runtime operations; it is not a generic Go library listing or a plain API framework entry.
What this skill actually does
Inputs and prerequisites: Go toolchain, tRPC-Agent-Go module, model provider credentials, optional MCP/A2A/AG-UI integrations.
Setup notes: Follow the upstream Go module setup from the tRPC-Agent-Go documentation, import the agent packages into a Go service, configure the model and tools, then run an example graph or runner with tracing enabled before production rollout.
Source and verification boundary: use https://trpc-group.github.io/trpc-agent-go/ as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Custom Agents workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.