PeepSo Friends adds a friend-graph layer to the PeepSo community plugin, storing each connection as a row in peepso_friends with user_id, friend_id, status (pending, accepted, blocked), and a created timestamp. Friend requests, accepted connections, and blocks all live in the same table with the status flag deciding which is which. The default PeepSo admin shows connections per profile but doesn't expose a cross-community queue for pending requests or a flat audit of who is blocking whom.

SleekView reads peepso_friends directly, joins users on both user_id and friend_id to surface both members' emails, and pivots the result into a flat connection table. Filter by status, sort by request date to clear stale pending requests, or group by user to spot accounts with unusually high outbound friend requests as a spam signal.

Inline edits route through PeepSo's friend model where exposed, so accepting or declining a request fires the standard notification hooks and updates cached friend counts. Direct table writes with conflict detection handle migrations, with the dual join making sure both sides of the connection stay in sync.

Social networks live and die on the quality of their friend graph. PeepSo Friends keeps the data simple because peepso_friends is a single edge table with status flags covering pending, accepted, and blocked connections. That makes everything queryable in principle.

The default admin only exposes it per profile, which is the wrong shape for cross-community moderation. A moderator wants the pending-request queue across every account so backlogs can be cleared in one pass. A trust-and-safety analyst wants the outbound-request leaderboard so high-volume requesters can be reviewed for spam-pattern signals.

A support agent triaging a complaint wants every connection a member has in one screen, including the blocks. SleekView's job is to compose those views directly against the same table PeepSo maintains, with writes routing through the friend model so notification hooks fire when they should. Same data, fewer screens, dramatically less profile-by-profile clicking for the people who run safety operations on live communities.