trpc-agent-go

The tRPC-Agent-Go MCP server exposes its agent orchestration capabilities over the MCP protocol, enabling other MCP clients to connect and coordinate multi-agent workflows within a Go-based intelligent agent framework. You can leverage its multi-agent orchestration, advanced memory, and tool integration features to compose complex pipelines, run agents in sequence or in parallel, and persist state across sessions. The MCP integration supports sending agent commands, tool invocations, and memory queries, as well as receiving structured results and telemetry. Typical usage involves launching the server (e.g., via a Docker image) and then connecting with an MCP client to register your tools, initialize agents, and start runs. You can use the built-in tooling to chain or parallelize sub-agents, harness persistent memory, integrate external tools, and observe runs with production telemetry.

Prerequisites:

• Docker installed and running
• Access to the Docker image trpc-agent-go-image (or build the image locally from the repository)

Step-by-step:

1. Pull the Docker image (if available): docker pull trpc-agent-go-image

2. Run the MCP server container: docker run -i -e TRPC_AGENT_GO_CONFIG=/path/to/config.yaml -e LOG_LEVEL=info trpc-agent-go-image

3. If you need to supply a custom config, mount it into the container: docker run -i
   -v /local/path/config.yaml:/path/to/config.yaml
   -e TRPC_AGENT_GO_CONFIG=/path/to/config.yaml
   -e LOG_LEVEL=info
   trpc-agent-go-image

4. Connect an MCP client to the running server using the MCP protocol and begin registering tools, agents, and runs. Refer to the project documentation for the exact MCP message formats supported by this server.

Tips:

• Ensure Docker image name matches the one published for this MCP server; if building locally, set the image tag accordingly.
• The server can accept configuration via environment variables or an external config file mounted into the container; use TRPC_AGENT_GO_CONFIG to point to the config path inside the container.
• Common issues include connectivity problems between the MCP client and server; verify network access and correct MCP port mappings if you customize ports.
• For production, enable telemetry and tracing by setting appropriate environment variables (e.g., LOG_LEVEL, TRACE_ENABLED) and ensure credentials (API keys) are provided securely.