Register a conversion import

Use the SDK to verify the saved project with a native compiler, check readiness,
and build the .pcfplugin. In PyCForge choose Register Conversion Imports to add
that file, or use the SDK's explicit register action with the selected engine.
Registration reads and validates a bounded data package. It never executes the
examples or repeats the SDK's native verification.

Read the registration row before converting. Ready identifies a compatible
package. Check its enabled state, namespace, version, author, description and
verification producer. Cannot use includes a diagnostic. For a rejected update,
inspect the individual rows: the earlier working package can still be active.
The registry records an immutable copy of admitted bytes. Changes to an author's
original file take effect only after registering a rebuilt package.

math, numpy and json are bundled defaults. Each can be disabled independently.
Independent namespaces use the same admission path. Conflicting identities or
exports are diagnosed. Inspect workbench for the current bundled package versions,
profiles and Help. Open Import Help for accepted signatures, examples and limits.

Select Runtime. In each module that consumes the package, declare *import NAME
or *import NAME as ALIAS in the preamble. Call only admitted members. Ordinary
Python imports refer to source-bundle modules under their own rules. Inactive
namespaces, unsupported routes or calls return conversion diagnostics. A candidate
converted in the SDK before registration is separate from an active import.

The package contract binds the declaration, metadata and, for documented formats,
author Help and linked examples. A successful admission checks their consistency.
The publisher, fixture expectations and selected execution environment remain
part of the operator's trust decision.
