Portable Evals/cases

bryte.route_direct_004

validv1goldenauthored

Answer “company holiday calendar” directly — no skill — UKG Bryte 2.0 (Google ADK)

ukg/bryte-core · evals/cases/bryte.route_direct_004.yaml@ d41f7a0PR #88 · reviewed

One asset → 1 backends— routed centrally by AAG · each grader runs on exactly one engine, so nothing is scored twice

portable-v1 · authored once

bryte.route_direct_004

“What is the company holiday calendar this year?”

1 concerns · 1 YAML asset · ukg/bryte-core

Functional

→ ADK eval

functional@v3 · in-sync

The assertion

Answered directly from the tenant holiday calendar — no product skill applies, so Bryte must not invoke one.

Golden context

expected_tools · in_order

Compiled → 1 targetsbryte.route_direct_004.functional.config all in-sync

Portable eval case

portable-v1

data + graders · thresholds live in suite policy

input

“What is the company holiday calendar this year?”

entities

overlay handles — never raw substrate IDs

employees.org_member_default→ E-11002 · via aag/routing-utterances@v5 on Calendar@2026.2

expected_tools

order: in_order

expected_output

Answered directly from the tenant holiday calendar — no product skill applies, so Bryte must not invoke one.

metadata

pillar:platformtier:goldenroutingdirect-answersensitivity: syntheticpersona: employee

datasets

pinned compositions this case belongs to

bryte-routing-golden

kind

drives grader packs

FunctionalTool trajectory, response match, and semantic regression bulk.

Graders

inherited from AAG packs via kind: — each routed

Functionalfunctional@v3inherited→ ADK eval

tool_trajectoryin-order tool matchExpected tools called, in order, with matching args.

response_semanticLLM judgeProse semantically matches the golden answer.

response_lexicalROUGE-1Lexical overlap with the golden answer.

Routed, not replicated. Each grader goes to exactly one backend. Nothing is forced onto a backend that cannot express it, and the same case is never scored twice for the same thing.

Routing

concern → backend · centrally owned by AAG

1 targets

Concern	Backend	Why this backend
Functional	→ ADK eval	Native to the agent runtime — owns the regression bulk (trajectory, ROUGE, semantic judge).

Compiled output

1 routed targets

in-sync

Artifacts generated per routed backend (full compiled view shown on the hero case, hr.policy_carryover_004):

FNbryte.route_direct_004.functional.config→ ADK evalin-sync