Explanation

Build a mental model before choosing commands or designing a repository:

• Why metadata as code? connects file-backed records to normal review and automation practices.
• The Mergeway repository model explains repositories, workspaces, entities, and records.
• The validation model explains the checks shared by Mergeway tools.
• Semantic diff versus file diff explains logical records, snapshots, and why line changes are not always data changes.
• MCP access and security explains read-only tools, entity scoping, and the surrounding trust boundaries.
• Editor architecture explains the LSP process, workspace ownership, in-memory overlays, and validation feedback loop.
• Schema design choices compares schema formats, record granularity, inheritance, unions, and identity strategies.
• Component architecture traces the shared repository core through each runnable interface.

The component overview shows how the released tools relate to the shared specification. A scaling guide is deferred until measured operational evidence can support useful thresholds and trade-offs; the site does not invent capacity guidance from implementation structure alone.