Every rating is done by its own AI agent running on Omnara. It investigates your account like an analyst, then writes the verdict.
You reply on ๐.Reply anything to the launch post to get rated, or tag a friend to rate them. We pull the account's profile and last ~200 posts, with their views, likes and replies.
An agent investigates.It has tools to read your profile, crunch your stats, pull your top, worst and most recent posts, and group your posts into topics. It decides what to look at, like an analyst would, and every step shows up in the report's log.
It scores you out of 10.Content, profile and originality are the agent's call. Consistency and engagement are computed from your numbers, so they don't change between runs. Your overall score is the average of all five.
You get the report.The agent replies on ๐ with your score card and a link to the full report: your archetype, best and worst post, what you post about, how your hits open, and the three fixes that would matter most.
Ask the judge.The agent sticks around after the report, so you can chat with it about the account.
Build your own
The judge is one Omnara agent profile with six custom tools: get_profile, get_stats, list_posts, list_all_posts, set_topics and submit_report. Our server answers each tool call, so the agent only ever sees your data through those tools.