Routing outcomes and failure modes

Routing can succeed, fail cleanly, or succeed in a degraded evidence state.

Successful selection

Success means:

• at least one candidate was eligible
• at least one candidate was scored
• chosen_endpoint_id is populated
• selection_reasons explain the winning choice

If more than one candidate was eligible, the decision also carries a fallback chain.

No eligible endpoint

This is the canonical routing failure shape:

• all candidates appear in eligibility
• each rejected candidate carries exclusion details
• chosen_endpoint_id is empty
• fallback_endpoint_ids is empty
• selection_reasons is empty

This is not an exception-shaped outcome. It is still a valid protocol artifact.

Policy conflict

Policy conflict usually appears as "all candidates rejected by policy," for example:

• every endpoint is denied by allow/deny rules
• remote routing is forbidden but no local candidate exists
• role-forbidden capabilities eliminate the remaining candidates

The protocol communicates this through candidate-level exclusions rather than through one special top-level error code.

Degraded selection with defaults

A router can still select an endpoint without observed performance evidence. In that case the decision will typically include:

• DEFAULT_PROFILE_USED
• used_measured: false

This is a degraded-evidence success, not a hard failure.

Insufficient evidence is usually not a rejection

In the reference router, lack of measured data is usually handled by:

• neutral default metric scores
• weight redistribution when an entire metric is unknown

That means "insufficient evidence" is typically represented as a weaker, more default-driven decision rather than as a hard exclusion.