Claude Code 多 agent 团队：Jev 路由 4800+ 微决策
用 Opus 5.5 做队长、三个 Haiku 5.5 并行干活，Jev 在底下处理「读哪个文件、用哪个模型、重试还是停」这类微决策，队长只在最后签字。
STOP MICROMANAGING YOUR F*CKING AGENTS Anthropic just dropped the cleanest agent team setup in Claude Code. Opus 5.5 leads the main session and builds the task DAG. Three Haiku 5.5 teammates self-claim work and talk to each other over a peer p2p bus. They run in parallel: → security audits JWT claims and finds a timing leak → perf-scout profiles redis locks and hits 2.8x faster → qa-engine runs the full test matrix (58/58 green) Atomic mutex DAG locks files so they never collide. JEV underneath resolves 4,800+ micro-forks (which file, which model, retry or stop) before anything reaches the lead. Opus only signs off at the end. 12/12 tasks done. Consensus approved. Zero regressions. You launch it like this: claude --model opus --teammate-mode auto The full tree > Opus 5.5 team lead builds the task DAG and coordinates > Haiku 5.5 x3 teammates - security / perf-scout / qa-engine > peer p2p bus for direct mailbox and self-claim tasks > atomic mutex DAG locks files with zero leader polling > JEV fork layer routes deterministic decisions > Opus signs off only after consensus and zero regressions Paste the tree + this prompt into Claude Code ↓ Rebuild my Claude Code setup around this tree: Run the main session on Opus 5.5 with effort xhigh and --teammate-mode auto. Spawn three Haiku 5.5 teammates: security, perf-scout, and qa-engine. Let them self-claim tasks from a shared DAG and communicate over a peer p2p bus. Route deterministic forks (which file, which model, retry or stop) through JEV. Use an atomic mutex DAG so teammates never collide on the same file. Opus only coordinates the plan and signs off at the end. Teammates handle the work. Show me every change as a diff first. No edits until I say go. save it, it’ll come in handy.
这条是引用着另一条写的：x.com/mirku21/status/2098836008468992471