Implement variable classification model and UI editing for literals, parameters, generated values, fixtures, secrets, and ignored data

[kalen-releasedgroup]

Summary

Give authors explicit control over how captured values should behave in generated tests.

Scope

• Persist variable/data strategy metadata for captured values.
• Support literal, scenario parameter, generated value, fixture reference, secret, and ignored-data modes.
• Reflect classification changes safely in previews and regeneration.

Requirement Links

• FR-INF-006, FR-AUTH-004
• Sections 7.3, 9.3, 12.3, 12.5

Dependencies

• Timeline editing and scenario version model

Exit Criteria

• Authors can classify all relevant values in the UI.
• Secret classifications remain masked in preview surfaces.
• Regeneration preserves approved data strategies.