Record the direct Agent Proxy Design gate verdict #164

New issue

Closed

opened 2026-07-31 07:52:10 +00:00 by coilyco-ops · 0 comments

coilyco-ops commented

2026-07-31 07:52:10 +00:00

Member

Parent: #163

Outcome

Commit the frozen direct-model Design evaluation evidence even though the v2 release gate fails. Keep #163 open for the required all-frontier-pass outcome.

Acceptance

The result records six frontier and six OSS cases sent directly through Agent Proxy logical routes, with no harness.
Governed Agent Proxy failures are preserved as model outcomes without retries.
The current Design pack digest and source revision validate.
Historical scorecard mode is removed and the current scorecard is regenerated.
The other seven role result files remain unchanged.
The declared test and smoke workflows pass.

Parent: #163 ## Outcome Commit the frozen direct-model Design evaluation evidence even though the v2 release gate fails. Keep #163 open for the required all-frontier-pass outcome. ## Acceptance * The result records six frontier and six OSS cases sent directly through Agent Proxy logical routes, with no harness. * Governed Agent Proxy failures are preserved as model outcomes without retries. * The current Design pack digest and source revision validate. * Historical scorecard mode is removed and the current scorecard is regenerated. * The other seven role result files remain unchanged. * The declared test and smoke workflows pass.

coilyco-ops referenced this issue from a commit

2026-07-31 07:53:09 +00:00

eval: record direct Design gate verdict, closes #164

coilyco-ops referenced this issue from a pull request that will close it,

2026-07-31 07:53:21 +00:00

eval: record direct Design gate verdict #165