Add gRPC task service with streaming results

[hyp3rd]

Summary

• define WorkerService proto with RegisterTasks and StreamResults endpoints
• implement gRPC server wiring worker Service to schedule tasks and stream results
• add generated protobuf and gRPC bindings

Testing

• buf lint
• pre-commit run --files grpc.go proto/worker/v1/worker.proto buf.yaml buf.gen.yaml pkg/worker/v1/worker.pb.go pkg/worker/v1/worker_grpc.pb.go go.mod go.sum (fails: go: no such tool "covdata")
• go test ./...

---

https://chatgpt.com/codex/tasks/task_e_68a0e3aadd8083309280653889c93de6

[Copilot]

Pull Request Overview

This PR adds gRPC service capabilities to the worker package by defining a WorkerService proto with task registration and result streaming endpoints. The implementation provides a gRPC server that wraps the existing worker Service to handle task scheduling and stream results back to clients.

• Defines protobuf messages and service for task registration and result streaming
• Implements gRPC server wrapper around existing worker Service
• Adds generated protobuf and gRPC bindings

Reviewed Changes

Copilot reviewed 6 out of 8 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
proto/worker/v1/worker.proto	Defines WorkerService with RegisterTasks and StreamResults RPCs
pkg/worker/v1/worker_grpc.pb.go	Generated gRPC service bindings and client/server interfaces
pkg/worker/v1/worker.pb.go	Generated protobuf message types and serialization code
grpc.go	Implements GRPCServer that wraps worker Service for gRPC endpoints
go.mod	Adds gRPC and protobuf dependencies
buf.yaml	Configures buf for protobuf management

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The word "pancis" should be "panics".

Suggested change

	// If the following call pancis, it indicates UnimplementedWorkerServiceServer was
	// If the following call panics, it indicates UnimplementedWorkerServiceServer was

[Copilot]

The task function always returns the payload regardless of context or arguments, which doesn't align with typical task execution patterns. Consider implementing actual task logic or storing the payload for later execution.

Suggested change

	task, err := NewTask(ctx, func(ctx context.Context, args ...any) (any, error) {
	task, err := NewTask(ctx, func(ctx context.Context, args ...any) (any, error) {
	if len(args) > 0 {
	return args[0], nil
	}

[Copilot]

[nitpick] The StreamResults method doesn't use the request parameter. Consider adding filtering options to the StreamResultsRequest to allow clients to specify which results they want to receive.

Suggested change

	Id: res.Task.ID.String(),
	// Assume req.Ids is a slice of task IDs to filter by. If empty, stream all results.
	idFilter := make(map[string]struct{})
	for _, id := range req.GetIds() {
	idFilter[id] = struct{}{}
	}
	for res := range s.svc.StreamResults() {
	taskID := res.Task.ID.String()
	// If filter is set, skip results not in the filter.
	if len(idFilter) > 0 {
	if _, ok := idFilter[taskID]; !ok {
	continue
	}
	}
	out := &workerpb.StreamResultsResponse{
	Id: taskID,