Editor Extensions

Vespera has a host-side extension registration API built around semantic editor commands and capabilities.

The important design rule is that extensions operate through supported editor operations instead of receiving raw pointers to scene vectors, renderer objects, CLR internals, or undo storage.

What the extension surface describes

The host registry can describe:

• extension identity and compatibility;
• semantic commands;
• whether commands mutate project state;
• undo/Play-mode capability metadata;
• discoverable capabilities shown through the editor/automation surface.

The Vespera core editor itself registers its semantic command surface through this mechanism, which keeps the public transaction model exercised by normal tooling.

Current boundary

The registry is not a general “drop any native DLL into a Plugins folder” loader. Arbitrary third-party native module loading is intentionally outside the current production workflow.

If you are building tooling today, prefer the semantic editor/automation interfaces documented under Automation / MCP instead of depending on engine-private structures.

Why this matters

A semantic command boundary gives tools a chance to preserve:

• validation;
• undo/redo behavior;
• stable IDs;
• Play/Edit mode rules;
• error reporting;
• future compatibility.

Direct mutation of internal containers would bypass those guarantees.