trails-discriminate-union

To identify discriminant needs for union-like branches in API schemas, analyze public Trails outputs across projections and surface responses to detect stable discriminants like kind, mode, or type. This process evaluates branch evidence and risk to determine if consumers can reliably distinguish branches.

Stable discriminants are properties like kind, mode, or type that allow consumers to reliably distinguish union branches in queryable shapes. You need them when exposing union-like public API outputs to agents or surfaces, ensuring schema stability and safe Warden enforcement.

Analyze Zod schemas and JSON Schema projections for branch evidence by reviewing Trails outputs to detect stable discriminants across tests and downstream surface behavior. This identifies whether schema changes are warranted or if advisory follow-ups are sufficient for governance.

Yes, this schema analysis tool works with public API surfaces and Warden enforcement by evaluating union-like branch outputs. It determines whether a stable discriminant exists, indicating when Warden enforcement is safe or when to raise advisory follow-ups and keep private unions internal.

You should not add a discriminant to a union schema when the analysis determines the issue is schema cleanup or advisory guidance rather than a stable discriminant requirement. In these cases, keep private unions internal and raise advisory follow-ups instead of enforcing Warden actions.

The best way to govern union-like branches exposed in public Trails outputs is to analyze branch evidence and risk across projections. This determines whether to add a discriminant, keep private unions internal, or raise advisory follow-ups based on authoritative source audits.