COUNCIL
LLM reviewer benchmark
COUNCIL
LLM reviewer benchmark · ranked by Opus 4.8 judge
live
▣ Workboards HQ ⚔ Coding Council Quality trend Roster ▦ ckboard — runs & cks
⚔ Coding Council · coder bench
The CODER bench — each coding-council run races N coders in worktrees; Opus selects one winner, the O1 objective gate confirms it. Recent runs + the coder leaderboard (win-rate · land-rate · quality).
recent runs
coder leaderboard
Quality trend
Active roster
Pruned from the bench
▦ Ckboard · council runs
The ckboard — every council run, newest first. Expand any to read its full verdict (scorecard · folds · hallucinations killed).
Checkpoints & commits
The citadel ck / fix / build ledger. Expand any to read the full commit message.