The AP2 primitives (checkout/payment mandates, Ed25519 sign/verify, the
x402 rail reference, attach-to-message) already existed, but the gateway
only *carried* mandates on the resulting task — it never verified them, so
ap2Verifications was never populated and a downstream paid path had no
trust signal.
Wire opt-in verification: set Options.AP2PublicKey (gateway) or
a2a.WithAP2PublicKey (embedded handler) and each mandate carried on a task
is verified (signature + task/context binding) with the result recorded in
task.AP2Verifications; the x402 settlement rail rides along for the paid
path. Off by default — mandates stay carried-but-unverified — so no payment
trust decision enters the default flow.
Adds a gateway integration test driving a real message/send that carries a
signed x402 payment mandate (verified, rail carried; tampered → surfaced as
unverified) plus a default-path test proving carry-only is unchanged.
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
Co-authored-by: Claude <noreply@anthropic.com>
The A2A gateway's push-notification flow (tasks/pushNotificationConfig/set
→ deliverPush) POSTed task state to a caller-supplied URL via the default
HTTP client, so an untrusted A2A caller could aim the gateway at internal
addresses (loopback, link-local cloud metadata, RFC1918) it would
otherwise never reach — a server-side request forgery vector (#4129).
Add a default SSRF-safe policy: only http/https callbacks whose host does
not resolve to a loopback, private, link-local, multicast, or unspecified
address. It's enforced when the config is set (caller gets a clear
rejection, nothing stored) and again at delivery, and the delivery client
re-checks the resolved IP at dial time so a name that passes validation
can't be rebound to an internal address before connect.
Operators that need a trusted in-cluster receiver set Options.AllowPushURL
(gateway) or a2a.WithPushURLPolicy (embedded handlers) to own the policy;
that path skips the built-in private-IP dial guard by design.
Tests cover blocked/allowed URLs, the dial-time guard, set-time rejection,
default-deny delivery, and the operator override.
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
Co-authored-by: Claude <noreply@anthropic.com>
* a2a: conform to external A2A clients — well-known path + spec SSE events
The A2A gateway interoperated go-micro-to-go-micro but a real external
client (ADK, LangGraph, a2a-SDK) would not:
- Discovery: served the Agent Card at /.well-known/agent.json, but A2A
0.3.0 discovers it at /.well-known/agent-card.json. Serve both, with
agent-card.json canonical and agent.json a legacy alias — per-agent,
per-skill, and at the single-agent top level.
- message/stream emitted repeated full Task snapshots. External SSE
clients parse by `kind` and stop on `final:true`; a Task snapshot has
neither, so they never terminate. Emit spec-shaped TaskArtifactUpdate
(append) chunks and close with a TaskStatusUpdate final:true. The
non-streaming and resubscribe paths also close with a terminal marker.
- A streaming error set both `result` and `error` in one JSON-RPC
response (strict clients reject it). Emit a failed status-update
instead — never result and error together.
Tests assert the canonical card path, the status-update/artifact-update
event shapes ending in final:true, and that no response carries both
result and error.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
* harness: update a2a-streaming to the spec-shaped stream events
The A2A gateway now emits artifact-update deltas and a terminal
status-update (final:true) instead of repeated full Task snapshots, so
the conformance harness must reassemble the answer from the append
artifact-update chunks and assert the final:true marker. This makes the
harness a stronger spec check rather than a snapshot-shape check.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
* agent: update a2a stream test to spec-shaped events
TestA2AStreamUsesAgentChatPathWithTools decoded the last SSE event as a
completed Task snapshot with artifacts. The gateway now closes the stream
with a status-update (final:true) and carries the answer as append
artifact-update deltas, so reassemble the answer from those deltas and
assert the terminal completed status-update instead.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
---------
Co-authored-by: Claude <noreply@anthropic.com>
Closes#4813. The stdio transport is the path an external MCP host (Claude
Desktop) uses, and it emitted broken output:
- tool results were `fmt.Sprintf("%v", decodedJSON)` → Go map-syntax
(`map[id:1 name:bob]`), not JSON. Now returned as JSON text.
- tool-execution failures were returned as JSON-RPC protocol errors; per the
MCP spec they must be a result with `isError:true` so the agent can read the
failure. Now they are (span/audit still record the error).
Both fixes are shared between stdio and websocket via a new `mcpToolResult`/
`mcpToolError` (dedupes the two transports). Added the missing stdio round-trip
tests (the package had zero) proving JSON output and the isError contract, using
an injected fake client; updated the websocket auth tests that asserted the old
protocol-error-on-tool-failure behavior.
Also fixes a pre-existing golangci-lint failure on master (unnecessary
`string(...)` conversion in grpcreflect.go from #4821) so the mcp package lints
clean — another one the required-checks gap let through.
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
Co-authored-by: Claude <noreply@anthropic.com>
* gateway/mcp: configurable initialize + named not-found
NewHandler takes HandlerOptions: WithServerInfo(name, version) and WithProtocolVersion set the initialize response so a product can brand the endpoint instead of advertising go-micro-mcp/2024-11-05. ManualResolver's not-found error now includes the tool name.
* gateway/mcp: configurable initialize + named not-found
NewHandler takes HandlerOptions: WithServerInfo(name, version) and WithProtocolVersion set the initialize response so a product can brand the endpoint instead of advertising go-micro-mcp/2024-11-05. ManualResolver's not-found error now includes the tool name.
* gateway/mcp: richer resolver Call + spec-correct HTTP handler
Resolver.Call now returns (*CallResult, error): a tool that ran but failed sets CallResult.IsError (returned as a tools/call result with isError:true, per the MCP spec), while a protocol/pre-check failure returns an error -- an *RPCError carries a specific JSON-RPC code. The HTTP handler also answers notifications/* (and id-less requests) with 204 and no body.
* gateway/mcp: richer resolver Call + spec-correct HTTP handler
Resolver.Call now returns (*CallResult, error): a tool that ran but failed sets CallResult.IsError (returned as a tools/call result with isError:true, per the MCP spec), while a protocol/pre-check failure returns an error -- an *RPCError carries a specific JSON-RPC code. The HTTP handler also answers notifications/* (and id-less requests) with 204 and no body.
* gateway/mcp: richer resolver Call + spec-correct HTTP handler
Resolver.Call now returns (*CallResult, error): a tool that ran but failed sets CallResult.IsError (returned as a tools/call result with isError:true, per the MCP spec), while a protocol/pre-check failure returns an error -- an *RPCError carries a specific JSON-RPC code. The HTTP handler also answers notifications/* (and id-less requests) with 204 and no body.
* gateway/mcp: swappable Resolver + mountable JSON-RPC HTTP handler
Adds a Resolver interface (Manual/Registry) so the tool source is
pluggable, and NewHandler(resolver) serving standard MCP JSON-RPC over
HTTP as an http.Handler you can mount on your own server. Neither
resolver exposes the internal store/broker tools. Additive — the
existing Serve() path is unchanged.
* gateway/mcp: swappable Resolver + mountable JSON-RPC HTTP handler
Adds a Resolver interface (Manual/Registry) so the tool source is
pluggable, and NewHandler(resolver) serving standard MCP JSON-RPC over
HTTP as an http.Handler you can mount on your own server. Neither
resolver exposes the internal store/broker tools. Additive — the
existing Serve() path is unchanged.
* gateway/mcp: swappable Resolver + mountable JSON-RPC HTTP handler
Adds a Resolver interface (Manual/Registry) so the tool source is
pluggable, and NewHandler(resolver) serving standard MCP JSON-RPC over
HTTP as an http.Handler you can mount on your own server. Neither
resolver exposes the internal store/broker tools. Additive — the
existing Serve() path is unchanged.
Fixes#2988. Brings 'golangci-lint run ./...' to zero issues (was ~373):
- errcheck: explicitly ignore fire-and-forget calls with '_ =' (and a small
errcheck.exclude-functions list for response writes — json Encoder.Encode,
http ResponseWriter.Write, fmt.Fprint*); genuine cases handled.
- unused: remove dead code (unexported decls and dead test helpers) and the
imports they orphaned.
- staticcheck: ST1005 error strings, ST1016 receiver names, S1000/S1017/S1019/
S1023 simplifications, SA4004/SA4006/SA4010 dead code, SA1021 net.IP.Equal,
SA6002 (store *[]byte in sync.Pool).
- govet: fix a context leak (lostcancel) in internal/util/mdns and move
t.Fatal/Fatalf out of goroutines (testinggoroutine) in tests.
- ineffassign, unconvert: mechanical fixes.
CI: the Lint workflow now runs a blocking full-tree 'golangci-lint run' on
pushes and PRs (dropped only-new-issues now that the tree is clean).
Verified: go build, go vet, test compilation, and unit tests for the
behaviourally-touched packages all pass.
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
Co-authored-by: Claude <noreply@anthropic.com>
* docs: compare Go Micro with Google ADK in the comparison guide
Adds a 'vs Agent Frameworks (Google ADK)' section: ADK builds an agent,
Go Micro builds the distributed system the agent lives in (agents are
services in the mesh). Covers the category difference, a feature table,
when to choose each, and MCP/A2A interoperability.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
* docs: replace ADK comparison slogan with concrete explanation
State plainly what each tool provides (ADK builds an agent process; Go Micro
builds the surrounding service mesh) instead of marketing phrasing.
* lint: apply golangci-lint autofixes; exclude ST1003 and demo errcheck
Mechanical, behaviour-preserving fixes applied by 'golangci-lint run --fix':
gofmt, misspell (US spelling), usestdlibvars (http.Method*/Status*), unconvert,
and the auto-fixable staticcheck simplifications (QF*, S1017/S1019/S1023/S1039).
Config: exclude ST1003 (remaining offenders are exported API renames, e.g.
web.Id, which would break compatibility) and skip errcheck for examples/ and
internal/harness/ (demo code where fire-and-forget is intentional).
Build and test compilation verified.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
* lint: WIP cleanup checkpoint (errcheck config + partial fixes)
Checkpoint of an in-progress golangci-lint cleanup (background pass). Builds
cleanly; lint is not yet zero. Follow-up commit will complete the cleanup and
switch CI to a blocking full-tree lint.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CmdEY7pYmV5zzwCjNJ4ykL
---------
Co-authored-by: Claude <noreply@anthropic.com>
* test(harness): read agent plan from the scoped store
The store-scoping change moved an agent's plan from the default table
key agent/{name}/plan to its own table (database "agent", table {name},
key "plan"). The plan-delegate harness tests still read the old key and
failed with 'not found'; read through store.Scope(mem, "agent", name)
like the agent does.
* docs: orient agents-first across README, landing, and docs overview
Lead with agents (then services and flows), surface MCP + A2A as the
interop story, and frame agents as services. Landing hero and feature
grid reordered agents-first with an A2A gateway card.
* v6: module path go-micro.dev/v6, TLS secure by default, NewService
Cut v6. Three breaking changes, bundled so the major bump is paid once:
- Module path go-micro.dev/v5 -> go-micro.dev/v6 across all imports + go.mod.
- TLS verification on by default (was off). MICRO_TLS_SECURE removed;
MICRO_TLS_INSECURE=true opts out for self-signed/dev.
- micro.NewService(name, opts...) is the canonical service constructor,
symmetric with NewAgent/NewFlow; micro.New kept as a deprecated alias;
the old name-less NewService(opts...) removed. Generators emit NewService.
Also ports the JWT auth token provider in-module (go-micro.dev/v6/auth/jwt/token
on golang-jwt/jwt/v5), dropping the v5-pinned github.com/micro/plugins/v5/auth/jwt
and the deprecated dgrijalva/jwt-go.
Docs/README/landing updated to v6 and @latest; v5->v6 migration guide added;
CHANGELOG cut as [6.0.0]. Blog posts left at their historical versions.
---------
Co-authored-by: Claude <noreply@anthropic.com>
Refactor the A2A handler into a reusable dispatcher + Invoke seam and
expose NewAgentHandler(card, invoke) + Card(). An agent now serves its
own A2A endpoint with AgentA2A(addr) / WithA2A — handling tasks
in-process (no RPC hop, no separate gateway). The gateway and embedded
agent share the same handler; the only difference is RPC vs in-process
invocation. Docs, README, and changelog cover both deployment modes.
Co-authored-by: Claude <noreply@anthropic.com>
* feat(a2a): Agent2Agent protocol gateway
Add gateway/a2a — exposes registered agents over the open A2A protocol so
agents on other frameworks can discover and call them. Agent Cards are
generated from registry metadata (the same way the MCP gateway derives
tools from service endpoints); incoming A2A tasks translate to the
agent's existing Agent.Chat RPC, so there's no per-agent code.
v1 is the synchronous JSON-RPC binding: message/send returns a completed
Task, tasks/get retrieves it, and Agent Cards are served for discovery;
streaming and push notifications are advertised as unsupported. Run with
'micro a2a serve' (cmd/micro/a2a). Tests cover card generation,
message/send, tasks/get, listing, and unknown-method errors.
* docs: A2A guide, README contents + A2A section, universe A2A check
- Add a Contents table of contents at the top of the README and an A2A
subsection under Building Agents.
- Add the Agent2Agent (A2A) guide and register it in the docs nav.
- Exercise the A2A gateway in the universe harness: the concierge agent
is reached over A2A (message/send -> Agent.Chat -> completed task).
* feat(a2a): outbound client — call external A2A agents
Add a2a.Client (Send/Card) so a Go Micro agent or flow can call an agent
on any framework by URL — the outbound counterpart to the gateway. Wired
in two places: flow.A2A(url) as a workflow step (the cross-framework
Dispatch), and agent delegate to an http(s) URL routes over A2A. The
universe harness now drives the gateway through the client, exercising
both directions. Tests cover client send/card and the round trip.
* docs: A2A both-directions — guide, README, changelog, blog #26
---------
Co-authored-by: Claude <noreply@anthropic.com>
* feat(mcp): advertise x402 payment requirements in the tool catalog
/mcp/tools now includes each priced tool's payment requirements (amount,
network, asset, payTo) when payments are enabled, so an agent can see the
cost before calling and choose by price — a shoppable catalog, the
foundation for a tool marketplace. Free tools carry no payment block; the
shared Tool struct is copied when pricing so it isn't mutated. Tests cover
priced/free tools and payments-disabled. Documented in the payments guide.
* feat(x402): consumer client with a spend budget (pay-and-retry)
Add x402.Client, the consumer counterpart to Middleware: it settles 402
challenges automatically via a pluggable Payer, up to a spend Budget. A
call that would exceed the budget is refused before any payment is made,
and spend accumulates across calls — the spend cap that keeps an
autonomous, paying caller in bounds. Tests cover pay-within-budget,
refuse-over-budget, budget accumulation, and free endpoints, end to end
against the server Middleware with a mock facilitator and payer. Guide
documents the consumer side; agent-level AgentMaxSpend is the next step.
* chore: gofmt gateway/mcp/benchmark_test.go (trailing newline)
---------
Co-authored-by: Claude <noreply@anthropic.com>
Follow-up to the merged x402 integration (#2964). Drop the commerce-y
'price' vocabulary for the protocol's own 'amount', and add per-tool
pricing as an operator concern (the way scopes/rate-limits are set at the
gateway).
- x402.Config: Price -> Amount (default), plus Amounts map for per-tool
overrides; AmountFor(tool) resolves per-tool -> default. Add a Require
primitive (per-request enforcement) and LoadConfig for an operator
config file.
- MCP gateway: enforce payment per-tool inside /mcp/call (where scopes
are enforced) using AmountFor, instead of a flat path-based middleware.
- CLI: --x402-price -> --x402-amount; add --x402-config (per-tool file)
to micro mcp serve and micro-mcp-gateway.
- docs: new Payments (x402) guide + nav + README section; blog/22
updated to Amount/Amounts and the config-file model.
Co-authored-by: Claude <noreply@anthropic.com>
Integrate the x402 payment protocol (HTTP 402) so a tool can require a
stablecoin payment and an agent can settle it — the next step after
autonomous agents (blog 21): agents that act, and pay.
- wrapper/x402: HTTP middleware enforcing the 402 challenge/verify flow,
with a pluggable Facilitator interface. Go Micro carries no chain or
crypto code — verification/settlement is delegated to a facilitator
(Coinbase CDP, Alchemy, self-hosted), so Base and Solana are just
different facilitators behind one interface. HTTPFacilitator default;
tests cover challenge / accept / reject via a mock facilitator.
- MCP gateway: optional Options.Payment gates /mcp/call (listing tools
and health stay free); off unless configured.
- micro mcp serve and micro-mcp-gateway: opt-in --x402-pay-to/-price/
-network/-facilitator flags (env vars on the standalone binary).
- blog/22 'Integrating x402: Payments for Agents'; README feature row.
Pricing is flat per call for now; richer models and an agent-side spend
cap (next to MaxSteps/ApproveTool) are follow-ups.
Co-authored-by: Claude <noreply@anthropic.com>
* feat: expose framework primitives via API gateway and MCP
Add registry, store, and broker as both HTTP routes and MCP tools
so AI agents and HTTP clients can inspect and operate the framework.
API gateway (/micro/* namespace):
GET /micro/registry List registered services
GET /micro/registry/{name} Describe a service
GET /micro/store List store keys
GET /micro/store/{key} Read a record
POST /micro/store/{key} Write a record
POST /micro/broker/{topic} Publish a message
MCP gateway (micro_* tool prefix):
micro_registry_list List services
micro_registry_get Describe a service
micro_store_list List keys
micro_store_read Read a record
micro_store_write Write a record
micro_broker_publish Publish a message
Framework tools use a Handler field on the MCP Tool struct for
direct dispatch (no RPC). Service tools continue to use RPC.
Rate limiters and circuit breakers are applied to framework
tools the same as service tools.
* fix: make framework internals opt-in on API and MCP gateways
Framework primitives (registry, broker, store) are now only
exposed when explicitly enabled:
API gateway: micro api --internal
MCP gateway: Options{Internal: true}
Off by default — user services are always exposed, framework
internals require the flag. Banner output only shows framework
routes when enabled.
* fix: always expose framework internals, gate by auth in production
Revert the --internal flag approach. Framework primitives (registry,
broker, store) are now always exposed:
- micro api: /micro/* routes always available (dev tool)
- MCP gateway: micro_* tools always registered. When Auth is
configured (production), they require micro:admin scope.
Without Auth (dev), they're open — same as all other tools.
This follows the existing pattern: micro run/api = dev (open),
micro server = production (auth + scopes). Framework internals
follow the same security model as user services.
Remove the Internal option from MCP Options. Remove --internal
flag from micro api.
Note: scope persistence depends on the store backend. The default
in-memory store does not survive restarts. Use MICRO_STORE=file
for persistent scopes in production.
* fix: correct DefaultStore comment — it's file-backed, not memory
* fix(server): don't recreate deleted admin user on restart
When the default admin account is deleted via the dashboard, set
a marker key (auth/.admin-deleted) in the store. On startup, skip
admin creation if the marker exists. This prevents the default
admin/micro credentials from reappearing after restart when the
user has intentionally removed them.
---------
Co-authored-by: Claude <noreply@anthropic.com>
* feat: add agent platform showcase and blog post
Add a complete platform example (Users, Posts, Comments, Mail) that
mirrors micro/blog, demonstrating how existing microservices become
AI-accessible through MCP with zero code changes.
Includes blog post "Your Microservices Are Already an AI Platform"
walking through real agent workflows: signup, content creation,
commenting, tagging, and cross-service messaging.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* refactor: rename handler types to drop redundant Service suffix
UserService → Users, PostService → Posts, CommentService → Comments,
MailService → Mail. Matches micro/blog naming convention.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* refactor: consolidate top-level directories, reduce framework bloat
Move internal/non-public packages behind internal/ or into their
parent packages where they belong:
- deploy/ → gateway/mcp/deploy/ (Helm charts belong with the gateway)
- profile/ → service/profile/ (preset plugin profiles are a service concern)
- scripts/ → internal/scripts/ (install script is not public API)
- test/ → internal/test/ (test harness is not public API)
- util/ → internal/util/ (internal helpers shouldn't be imported externally)
Also fixes CLAUDE.md merge conflict markers and updates project
structure documentation.
All import paths updated. Build and tests pass.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* refactor: redesign model package to match framework conventions
Rename model.Database interface to model.Model (consistent with
client.Client, server.Server, store.Store). Remove generics in
favor of interface{}-based API with reflection.
Key changes:
- model.Model interface: Register once, CRUD infers table from type
- DefaultModel + NewModel() + package-level convenience functions
- Schema registered via Register(&User{}), no per-call schema passing
- Memory implementation as default (in model package, like store)
- memory/sqlite/postgres backends updated for new interface
- protoc-gen-micro generates RegisterXModel() instead of generic factory
- All docs, blog, and README updated
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
---------
Co-authored-by: Claude <noreply@anthropic.com>
* docs: update all four documentation guides and mark Q2 complete
- ai-native-services: add WithMCP one-liner, standalone gateway,
WebSocket client example, and OpenTelemetry observability section
- mcp-security: add OTel distributed tracing, WebSocket authentication
(connection-level and per-message), DeniedReason audit field
- tool-descriptions: add manual overrides with WithEndpointDocs and
export formats section
- agent-patterns: add LangChain/LlamaIndex SDK pattern and standalone
gateway production pattern with Docker example
- Update roadmap: mark Q2 documentation as complete, Q2 at 100%
- Update status: reflect all recent completions, shift priorities
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add agent demo example and blog post
Add examples/agent-demo with a multi-service project management app
(projects, tasks, team) that demonstrates AI agents interacting with
Go Micro services through MCP. Includes seed data and example prompts.
Add blog post 4 "Agents Meet Microservices: A Hands-On Demo" walking
through the example code and showing cross-service agent workflows.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: enable multiple services in a single binary
Remove global state mutations from service and cmd option functions so
that configuring one service no longer overwrites another's settings.
Key changes:
- service/options.go: remove all DefaultXxx global writes from option
functions; newOptions() now creates fresh Server, Client, Store, and
Cache per service while sharing Registry, Broker, and Transport
- cmd/cmd.go: newCmd() uses local copies instead of pointers to package
globals; Before() no longer mutates DefaultXxx vars
- cmd/options.go: remove global mutations from all option functions
- service/service.go: export ServiceImpl type for cross-package use
- service/group.go: new Group type for multi-service lifecycle
- micro.go: add Start/Stop to Service interface, expose Group and
NewGroup convenience function
- examples/multi-service: working example with two services
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* docs: highlight multi-service binary support
Add multi-service section to README with code example, update features
list, add to examples index, and note in status summary.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: unify service API and clean up developer experience
- Unified service creation: micro.New("name", opts...) as canonical API
- Clean handler registration: service.Handle(handler, opts...) accepts
server.HandlerOption args directly, no need to reach through Server()
- Unexported serviceImpl: users interact through Service interface only
- Service groups use Service interface (not concrete type)
- Fixed Stop() to properly propagate BeforeStop/AfterStop errors
- Fixed store init: error-level log instead of fatal on init failure
- Updated all examples to use consistent patterns
- Updated README, getting-started, MCP docs, and guides
- Added blog post about the DX cleanup
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* fix: add blog post 5 to blog index
Blog post 5 (Developer Experience Cleanup) existed as a file but was
missing from the blog index page.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: make micro new generate MCP-enabled services by default
- main.go template includes mcp.WithMCP(":3001") by default
- Handler template has agent-friendly doc comments with @example tags
- Proto template has descriptive field comments
- README includes MCP usage, Claude Code config, and tool description tips
- Makefile adds mcp-tools, mcp-test, mcp-serve targets
- go.mod updated to Go 1.22
- Added --no-mcp flag to opt out of MCP integration
- Post-create output shows MCP endpoint URLs
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* docs: add MCP migration guide and troubleshooting guide
- Migration guide: 3 approaches to add MCP to existing services
(WithMCP one-liner, standalone gateway, CLI)
- Troubleshooting guide: common issues with agents, WebSocket,
Claude Code, auth, rate limiting, and performance
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* refactor: rename model/ package to ai/ for AI model providers
The model/ package name conflicted with the conventional use of "model"
for data models. Renamed to ai/ which better describes the package's
purpose (AI provider abstraction for Anthropic, OpenAI, etc.) and frees
up model/ for future data model layer use.
- Rename model/ → ai/ with package name change
- Update all Go imports from go-micro.dev/v5/model to go-micro.dev/v5/ai
- Update cmd/micro/server/server.go references (model.X → ai.X)
- Update all documentation and roadmap references
- All tests pass, CLI builds successfully
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add model package for typed data access with CRUD and queries
New model/ package provides a typed data model layer using Go generics.
Supports structured CRUD operations, WHERE filters, ordering, pagination,
and automatic schema creation from struct tags.
Three backends:
- memory: in-memory for development and testing
- sqlite: embedded SQL for dev and single-node production
- postgres: full PostgreSQL for production deployments
Key features:
- Generic Model[T] with Create/Read/Update/Delete/List/Count
- Query builder: Where(), WhereOp(), OrderAsc/Desc(), Limit(), Offset()
- Struct tags: model:"key" for primary key, model:"index" for indexes
- Auto table creation from struct schema
- 19 tests passing across memory and sqlite backends
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add model code generation to protoc-gen-micro
Extend the micro plugin to generate model structs from proto messages
annotated with // @model. Generated alongside client/server code in
the same .pb.micro.go file.
For a proto message like:
// @model
message User { string id = 1; string name = 2; }
Generates:
- UserModel struct with model:"key" and json tags
- NewUserModel(db) factory returning *model.Model[UserModel]
- UserModelFromProto(*User) *UserModel converter
- (*UserModel).ToProto() *User converter
Supports @model(table=custom_table, key=custom_field) options.
Adds GetComments() to generator for plugin comment inspection.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add Model() to Service interface for Client/Server/Model trifecta
Every service now exposes Client(), Server(), and Model() — call services,
handle requests, and save/query data from the same interface. Includes
README docs, blog post, and a full model guide on the docs site.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add Helm chart for MCP gateway Kubernetes deployment
Adds official Helm chart at deploy/helm/mcp-gateway/ with:
- Deployment, Service, ServiceAccount templates
- HPA for auto-scaling based on CPU/memory
- Ingress with TLS support
- Configurable registry (consul, etcd, mdns), rate limiting,
JWT auth, audit logging, and per-tool scopes
- Security context (non-root, read-only rootfs, drop all caps)
- NOTES.txt with post-install connection instructions
Updates roadmap and status docs to reflect Helm Charts as delivered.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* docs: add Helm chart entry to changelog
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add per-tool circuit breakers to MCP gateway
Protects downstream services from cascading failures. When a tool's
RPC calls fail repeatedly, the circuit opens and rejects requests
immediately until the service recovers (half-open probe pattern).
- CircuitBreakerConfig with MaxFailures, Timeout, MaxHalfOpen
- Per-tool breakers created during service discovery
- Integrated into HTTP call path with 503 response when open
- Records success/failure after each RPC call
- --circuit-breaker and --circuit-breaker-timeout CLI flags
- 8 unit tests covering all state transitions
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
---------
Co-authored-by: Claude <noreply@anthropic.com>
* docs: update all four documentation guides and mark Q2 complete
- ai-native-services: add WithMCP one-liner, standalone gateway,
WebSocket client example, and OpenTelemetry observability section
- mcp-security: add OTel distributed tracing, WebSocket authentication
(connection-level and per-message), DeniedReason audit field
- tool-descriptions: add manual overrides with WithEndpointDocs and
export formats section
- agent-patterns: add LangChain/LlamaIndex SDK pattern and standalone
gateway production pattern with Docker example
- Update roadmap: mark Q2 documentation as complete, Q2 at 100%
- Update status: reflect all recent completions, shift priorities
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add agent demo example and blog post
Add examples/agent-demo with a multi-service project management app
(projects, tasks, team) that demonstrates AI agents interacting with
Go Micro services through MCP. Includes seed data and example prompts.
Add blog post 4 "Agents Meet Microservices: A Hands-On Demo" walking
through the example code and showing cross-service agent workflows.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: enable multiple services in a single binary
Remove global state mutations from service and cmd option functions so
that configuring one service no longer overwrites another's settings.
Key changes:
- service/options.go: remove all DefaultXxx global writes from option
functions; newOptions() now creates fresh Server, Client, Store, and
Cache per service while sharing Registry, Broker, and Transport
- cmd/cmd.go: newCmd() uses local copies instead of pointers to package
globals; Before() no longer mutates DefaultXxx vars
- cmd/options.go: remove global mutations from all option functions
- service/service.go: export ServiceImpl type for cross-package use
- service/group.go: new Group type for multi-service lifecycle
- micro.go: add Start/Stop to Service interface, expose Group and
NewGroup convenience function
- examples/multi-service: working example with two services
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* docs: highlight multi-service binary support
Add multi-service section to README with code example, update features
list, add to examples index, and note in status summary.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: unify service API and clean up developer experience
- Unified service creation: micro.New("name", opts...) as canonical API
- Clean handler registration: service.Handle(handler, opts...) accepts
server.HandlerOption args directly, no need to reach through Server()
- Unexported serviceImpl: users interact through Service interface only
- Service groups use Service interface (not concrete type)
- Fixed Stop() to properly propagate BeforeStop/AfterStop errors
- Fixed store init: error-level log instead of fatal on init failure
- Updated all examples to use consistent patterns
- Updated README, getting-started, MCP docs, and guides
- Added blog post about the DX cleanup
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
---------
Co-authored-by: Claude <noreply@anthropic.com>
* feat: add LlamaIndex SDK for Go Micro services
Add LlamaIndex integration package that enables LlamaIndex agents to
discover and call Go Micro microservices through the MCP gateway.
Follows the same pattern as the existing LangChain SDK.
- GoMicroToolkit with from_gateway() factory and tool filtering
- FunctionTool integration via llama_index.core.tools
- Auth support, error handling, and retry configuration
- Examples for basic agent and RAG + microservices workflows
- Unit tests with mocked gateway responses
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* docs: update status for OTel, WebSocket, and LlamaIndex SDK completion
Reflect recently completed work in roadmap and status documents:
- Q2 progress: 85% -> 95% (WebSocket, LlamaIndex SDK done)
- Q3 progress: 40% -> 50% (OpenTelemetry integration done)
- Transports: 2 -> 3 (added WebSocket)
- Agent SDKs: 1 -> 2 (added LlamaIndex)
- Test coverage: 568 -> 1,000+ lines
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add WithMCP convenience option, improve startup banner, and blog post
- Add mcp.WithMCP(":3000") service option for one-line MCP setup
- Improve `micro run` startup banner to show Agent playground, MCP
tools, and WebSocket endpoints prominently
- Add blog post: "Building the AI-Native Future of Go Micro with Claude"
covering WebSocket transport, OTel integration, LlamaIndex SDK, and
Anthropic's Claude Max sponsorship
- Update blog index and navigation links
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Update docs and roadmap to March 2026 with focus priorities
- ROADMAP.md: Updated from Nov 2025 to reflect Q1 completions and current state
- ROADMAP_2026.md: Updated status to March 2026, added model package as delivered
- CURRENT_STATUS_SUMMARY.md: Rewrote with March 2026 status and clear next priorities
- PROJECT_STATUS_2026.md: Added model package section, updated recommendations
- Website roadmap: Updated Q3 security status and timestamps
Key focus areas identified: documentation guides, multi-protocol MCP,
LlamaIndex SDK, and OpenTelemetry integration.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* Add CLAUDE.md and four documentation guides to fill doc gaps
- CLAUDE.md: Project guide with structure, build commands, and priorities
- ai-native-services.md: End-to-end tutorial building an MCP-enabled task service
- mcp-security.md: Production security guide (auth, scopes, rate limiting, audit)
- tool-descriptions.md: Best practices for writing Go comments that help agents
- agent-patterns.md: Six integration patterns from single-agent to event-driven
- Updated docs index with new "AI & Agents" section linking all four guides
These were the highest priority gaps identified in the roadmap analysis:
the framework has solid features that were under-documented.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
* feat: add OpenTelemetry tracing to MCP gateway
Integrate OpenTelemetry spans into the MCP gateway for both HTTP and
stdio transports. Each tool call now creates a server span with rich
attributes (tool name, account ID, auth outcome, transport type).
Trace context is propagated to downstream RPC calls via metadata,
enabling end-to-end distributed tracing through Jaeger, Grafana, etc.
New files:
- gateway/mcp/otel.go: Span creation, attribute constants, metadata carrier
- gateway/mcp/otel_test.go: 8 tests covering span creation, auth denied/allowed,
rate limiting, trace propagation, noop provider, and missing token
Changes:
- Options.TraceProvider: Optional trace.TracerProvider field
- handleCallTool (HTTP): Creates OTel spans with auth/rate-limit attributes
- handleToolsCall (stdio): Same instrumentation for stdio transport
- go.mod: Added go.opentelemetry.io/otel/sdk v1.35.0 (test dependency)
The existing MCP trace ID (UUID) is preserved for backward compatibility
and recorded as a span attribute alongside the W3C trace context.
https://claude.ai/code/session_01GkduEhcrqcG45rdfYh8dAc
---------
Co-authored-by: Claude <noreply@anthropic.com>
* Initial plan
* Add MCP per-tool scopes, tracing, rate limiting, and audit logging
- Add Scopes field to Tool struct for per-tool scope requirements
- Add Auth (auth.Auth) integration to Options for token inspection
- Add trace ID generation (UUID) propagated via metadata to downstream RPCs
- Add per-tool rate limiting with configurable requests/sec and burst
- Add AuditFunc callback for immutable tool-call audit records
- Extract tool scopes from registry endpoint metadata ("scopes" key)
- Update both HTTP and stdio transports with auth/trace/rate/audit
- Add comprehensive tests for all new functionality
Co-authored-by: asim <17530+asim@users.noreply.github.com>
* Revert unrelated example go.mod changes
Co-authored-by: asim <17530+asim@users.noreply.github.com>
* Remove auto-generated example go.sum files
Co-authored-by: asim <17530+asim@users.noreply.github.com>
* Add WithEndpointScopes helper, gateway-level ToolScopes, and documentation
- Add server.WithEndpointScopes() for declaring per-endpoint auth scopes at
handler registration time
- Add mcp.Options.ToolScopes for gateway-level scope overrides without
changing individual services
- Update documented example to show WithEndpointScopes usage
- Update examples/mcp/README.md with scopes, tracing, and rate-limiting docs
- Update gateway/mcp/DOCUMENTATION.md with scopes section and FAQ
- Add tests for both new features
Co-authored-by: asim <17530+asim@users.noreply.github.com>
* Fix ToolScopes doc comment: clarify override (not merge) semantics
Co-authored-by: asim <17530+asim@users.noreply.github.com>
* Revert unrelated example go.mod/go.sum changes
Co-authored-by: asim <17530+asim@users.noreply.github.com>
* Rename ToolScopes to Scopes in MCP Options
The field name "Scopes" is more universal and consistent with how
auth scopes are used throughout go-micro. Updated all code references,
tests, and documentation.
Co-authored-by: asim <17530+asim@users.noreply.github.com>
* MCP gateway: add per-tool scopes, tracing, rate limiting, and audit logging
Co-authored-by: asim <17530+asim@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: asim <17530+asim@users.noreply.github.com>
Implement Q2 2026 roadmap items for AI-native microservices:
MCP stdio transport:
- JSON-RPC 2.0 over stdio for Claude Code integration
- Methods: initialize, tools/list, tools/call
- Auto-detection: stdio (no address) vs HTTP/SSE (with address)
micro mcp command:
- 'micro mcp serve' - start MCP server (stdio or HTTP)
- 'micro mcp list' - list available tools
- 'micro mcp test' - test a tool (placeholder)
- Enables Claude Code users to add microservices as tools
Gateway refactor:
- Created gateway/api package (reusable, 150 lines)
- Moved gateway logic from cmd/micro/server/gateway.go
- HandlerRegistrar pattern for flexibility
- cmd/micro/server/gateway.go now compatibility wrapper (72 lines)
- 50% code reduction, better separation of concerns
- Library users can now use gateway in custom apps
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Major Features:
- Unified gateway architecture (micro run + micro server use same code)
- MCP (Model Context Protocol) integration as library package
- AI-accessible microservices with 3 lines of code
Gateway Unification:
- Created reusable gateway module (cmd/micro/server/gateway.go)
- Updated micro run to use unified gateway (removed duplicate code)
- Conditional authentication (disabled in dev, required in prod)
- Reduced code duplication, simplified maintenance
MCP Integration:
- New library package: gateway/mcp
- Automatic service discovery → MCP tools
- HTTP/SSE transport support (stdio coming soon)
- Works for both library users and CLI users
- CLI flags: --mcp-address for micro run and micro server
Documentation:
- ADR-010: Unified Gateway Architecture
- CLI & Gateway Guide for users
- MCP Gateway README and examples
- Blog post: Making Your Microservices AI-Native with MCP
Breaking Changes: None (fully backward compatible)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Asim Aslam