HEAD 19588c0a04; both proof servers stopped. Invalid options must fail before trying the stopped server. $ t3 thread start . "Reply cli ok" --provider codex --model gpt-6.1-sol --option reasoningEffort=high --option reasoningEffort=low --wait --json --option reasoningEffort is given more than once. Exit: 1 $ t3 thread start . "Reply cli ok" --option reasoningEffort=high --wait --json --option needs --provider and --model. Exit: 1