https://www.oceanheart.ai/tags/multi-model/ Recent content in Multi-Model on oceanheart.ai Hugo en-gb Mon, 30 Mar 2026 00:00:00 +0000 https://www.oceanheart.ai/projects/sortie/ Mon, 30 Mar 2026 00:00:00 +0000 https://www.oceanheart.ai/projects/sortie/ <a class="claude-coach" href="https://claude.ai/new?q=Read&#43;https%3A%2F%2Fwww.oceanheart.ai%2Fprojects%2Fsortie%2F&#43;and&#43;discuss&#43;adversarial&#43;multi-model&#43;code&#43;review&#43;with&#43;me%3A&#43;convergence&#43;analysis%2C&#43;severity&#43;gating%2C&#43;and&#43;where&#43;single-reviewer&#43;blind&#43;spots&#43;survive." target="_blank" rel="noopener noreferrer" aria-label="Open a prepared Claude conversation"> <span class="claude-coach-mark" aria-hidden="true">&#10035;</span> <span class="claude-coach-copy"> <span class="claude-coach-eyebrow">Claude coach</span> <strong class="claude-coach-title">Interrogate this architecture</strong> <span class="claude-coach-description">Open a Claude conversation primed to discuss adversarial multi-model review: convergence, severity gating, and surviving blind spots.</span> <span class="claude-coach-action">Discuss it with Claude <span aria-hidden="true">&#8599;</span></span> </span> </a> <h2 id="what-it-is">What it is</h2> <p>An async adversarial code review system for multi-agent development workflows. Sortie runs a configurable roster of language models (Claude, Codex, Gemini) against a worker&rsquo;s diff in parallel, synthesises their findings through a 4th-model debrief invocation, triages by severity, and gates worktree merges. All findings, dispositions, and cost data are captured in a structured ledger.</p>