.Breeze v2 is a reasonably large step ahead.
Breeze began as a Go internet framework, however with v2 I’m pushing it towards one thing broader: a Go-native runtime with built-in developer tooling, observability and AI integration.
The three largest additions on this launch are Fleet, MCP and JSON-RPC.
Fleet
Distributed programs are simple to construct.
Understanding what occurred to a request throughout a number of providers is the exhausting half.
Breeze Fleet provides distributed tracing with W3C traceparent, permitting hint context to journey throughout providers and providing you with a strategy to comply with a request by way of the system.
The fascinating half is that Fleet isn’t utterly remoted from the remainder of Breeze.
It may be inspected by way of the MCP tooling as effectively.
MCP
MCP might be the largest architectural addition in v2.
Breeze helps MCP at two completely different ranges.
The primary is the CLI / generator facet.
An AI agent can work together with the Breeze improvement atmosphere by way of MCP to know a challenge, generate code, create plans and changesets, confirm modifications, diagnose issues, examine Fleet and extra.
The second is in-app MCP.
A working Breeze software can expose its personal capabilities as MCP instruments:
app.MCPTool("getUser", "Get a person by ID", handler)
This implies MCP isn’t solely a development-time function.
Your precise software can change into one thing an AI agent can uncover and work together with.
That’s a really completely different mannequin from merely placing an AI chatbot on prime of an API.
JSON-RPC 2.0
Beneath the MCP layer, Breeze v2 additionally introduces a correct JSON-RPC 2.0 implementation.
It contains:
This provides Breeze a protocol layer that can be utilized independently of MCP.
Placing it collectively
The fascinating factor isn’t actually any considered one of these options.
It’s how they join:
AI Agent
│
MCP
┌─────┴─────┐
│ │
CLI / Generator Operating App
│ │
└─────┬─────┘
│
JSON-RPC
│
Breeze
│
┌─────┴─────┐
│ │
Workflows Occasions
│ │
└─────┬─────┘
│
Fleet
│
Distributed Tracing
The aim with v2 is to not bolt AI onto a standard internet framework.
I’m making an attempt to make the runtime, developer tooling, protocols, observability and AI integration a part of the identical Go ecosystem.
Breeze v2 continues to be evolving, however that is the course.
Go Native. AI Native.
GitHub: GitHub – nelthaarion/breeze: Breeze — Excessive-Efficiency Golang Internet Framework · GitHub
#Golang #Go #MCP #JSONRPC #DistributedSystems #Observability #OpenSource #DevTools #Backend
