Data workflow

Data editing with JSON Patch

Data editing should be graph-owned and proof-validated. The Inspector can show a focused JSON tree, where-used rows, patch preview, refusal reasons, and rendered effects while the server remains the authority.

Patch proposal

Data changes start as explicit operations.

Replace scalar, insert item, or update object field proposals reference a data document, JSON pointer, old value digest, and expected affected fragments.

Rendered effect

Where-used makes impact visible.

The Inspector should show which templates and rendered ranges consume the value before the write is committed.

Stage	Required proof	Failure path
Locate value	Data document digest and pointer.	Reload data map.
Preview patch	Old value digest and type.	Reject stale data.
Validate effects	Where-used fragments and render context.	Refresh or refuse visibly.
Commit	Source/data after digest and journal row.	Mark target stale if effect realization fails.

Data-oriented fixtures shell

hypermake explain rendered/public/data/data-root.html
hypermake affected-by data/large-items-committed/006zjLluByRh.json
hypermake diagnostics src/data/data-root.html