Use this skill when the user is trying to force deeper computation through repeated rollouts or Prime Gauss style recursive prompting. It preserves the useful aspects of repeated trials, prior memory, fresh information, and explicit marks while discarding the assumption that the loop proves certainty. The ledger contract records trial counts, candidates, decision marks, coherence marks, and promotion gate results in JSONL or Markdown formats. Reach for it when you need to structure high-dimensional search, stochastic optimization, or ensemble comparison with a visible evidence trail.