Files
gbrain/test/e2e
Garry TanandClaude Fable 5 ea041756ae test(e2e): align cross-modal e2e stubs with the new DEFAULT_SLOTS models
The runner now dispatches to openai:gpt-5.2 / anthropic:claude-opus-4-8 /
google:gemini-3-pro; the e2e chat stubs still keyed on the old defaults, so
every verdict collapsed to inconclusive (4 failures in
test/e2e/cross-modal-eval.test.ts).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-22 11:16:58 -07:00
..