MCP Context

Who is this for? Stakeholders deciding whether to self-host Rosetta MCP — the business case, data-handling responsibilities, and opt-in features that come with it.

When should I read this? After Context, only if you’re evaluating the self-hosted MCP path. Most teams don’t need this — plugins are primary and cover the default case.

What this document is. Business context, data-handling responsibilities, and opt-in capabilities specific to self-hosted Rosetta MCP deployments. No technical implementation details — see MCP Architecture for internals, Deployment for setup.

What this document is not. Not the general business case for Rosetta (see Context), not architecture, not a deployment guide.

When You Need This

Self-hosted MCP is optional. Most organizations run Rosetta entirely through plugins and never need this. Consider self-hosting Rosetta MCP only if you specifically need:

If none of these apply, use Plugins instead.

Opt-In Features and Data Responsibility

Rosetta itself does not store any project data — that principle doesn’t change with MCP. A self-hosted MCP deployment can additionally opt into extra features; once enabled, your infrastructure stores the resulting data and you are responsible for it:

None of these are enabled by default, and none exist in plugin mode — there is no infrastructure to store data in, so there’s nothing to review.

Value Delivery Specific to Self-Hosted MCP

The adoption tracking, per-feature usage visibility, and transparent usage metrics mentioned in Context — Value Delivery for Directors and VPs come from this opt-in analytics stack. Plugin-only deployments don’t have them, and don’t need any data-residency review to avoid them.