[EPIC] Production Remote-Host Operations

## Problem statement

Remote Azure MCP deployments need more than a successful installation. Operators require repeatable configuration, capacity validation, resource-consumption visibility, request controls, and actionable connection diagnostics. Today those concerns are tracked independently, which makes production readiness difficult to assess across App Service, Azure Functions, containers, and Kubernetes environments.

## Vision

Remote Azure MCP deployments are configurable, observable, scalable, and diagnosable using supported operational controls and documented production baselines.

## Current open work

- [ ] #1570, automate remote-server load testing
- [ ] #1567, provide a resource-consumption telemetry dashboard
- [ ] #1101, implement outbound rate limiting
- [ ] #975, consolidate environment variables using .NET options
- [ ] #966, improve HTTP connection termination diagnostics

## Completed foundation

- #1736 delivered easier App Service deployment guidance and templates
- #1659 added multi-architecture Docker image support
- #1429 resolved Kubernetes HTTP transport deployment failures
- #1100 established CORS restriction and customization behavior

## Goals (in scope)

- Define production load and resource-consumption baselines
- Provide consistent typed configuration across hosting environments
- Add request and dependency controls that protect shared deployments
- Surface actionable operational telemetry and connection diagnostics
- Document supported production topology and scaling expectations

## Non-goals (out of scope)

- Owning MCP protocol conformance, which belongs to the stateless protocol epic
- Owning one-click deployment UX, which belongs to the MCP Hub self-hosting epic
- Replacing hosting-platform monitoring or autoscaling capabilities

## Success criteria

- [ ] Repeatable load tests cover representative remote workloads
- [ ] Operators can observe capacity, resource consumption, failures, and connection lifecycle events
- [ ] Configuration is consistent and validated across supported hosts
- [ ] Rate limits and dependency controls fail explicitly and predictably
- [ ] Production deployment and troubleshooting guidance reflects validated behavior

## Dependencies

- Supported remote hosting architectures and deployment templates
- Azure Monitor and application telemetry infrastructure
- MCP transport and authentication behavior
- Representative load profiles and production capacity targets

Begin by reviewing the five open work items: #1570, #1567, #1101, #975, and #966, along with the completed foundation issues listed in the epic. Treat the success criteria as the definition of done: validated load tests, observable capacity and failures, consistent configuration, predictable controls, and updated production guidance.

Written by the indexing model from the issue text.