If higher-credibility users end up concentrated in one ideological group, and lower-credibility in another, how should a system like this handle it?

Right now the design doesn’t attempt to enforce “balance.” Instead, it exposes the credibility scores transparently, lets users set their own filters, and tracks credibility at multiple levels (user, post, and domain). But I’d love thoughts from this community on whether that’s enough — or if more guardrails are needed.

Have you thought of how you might handle people using the +/- options as an agree/disagree ranking? I’d the site is small, it may be self-policing, but as it grows it can be a problem. Reddit ran into this. The upvote/downvote was supposed to be about if something contributed to the conversation, regardless of your opinion on it, but that eventually broke down. While giving the next on the button itself may help, there is still the possibility of abuse there.

Will votes +/- be weighted by the credibility score of the voter?

Another issue Reddit ran into was selling of accounts. Someone could build an account with a high credibility score and sell it to someone looking to spread misinformation.