Hi. I would like to only register my different prompts with customs labels for my combination of agents/model so that I manually can annotate them and assign the pass or no pass status. Without the actual run against any models API. I only want to make sure that the prompt is in the database and gives me the option to evaluate/mark/rate them, but the actual evaluation is done somewhere else, like in** claude code** or in codex. I launched

promptfoo eval

And immediately cancelled it. get the prompt in the database but there are no options to manually annotate them or rank them. Basically I need the evaluation metrics even if the problem didn't run - #evals https://cdn.discordapp.com/attachments/1429949929895493723/1429949931329683586/image.png?ex=68f80026&is=68f6aea6&hm=baba382cc34013385292eccf32ab9408222c2a47ff0715eba7e58f97ba05d31d& https://cdn.discordapp.com/attachments/1429949929895493723/1429949932080599221/image.png?ex=68f80026&is=68f6aea6&hm=ed959dc3f0dc443702e9aac5887b6df4b4eaccfbd2947aee4ff61e69fa5edff6&

Does exclusively using the

echo

provider work for you?

Hi. I'm very new to the platform. Could you please elaborate on this echo provider or link some docs? I don't even know what it is.