arenalisted

# Arena Fan out N parallel attempts at the same task. Read every candidate end to end. Pick the strongest as the base. Graft the best ideas from the others into it. Verify the synthesized result. ## Start Open a todolist with one entry per phase before launching anything. The arena runs autonomously and the list keeps phases from silently disappearing. 1. Frame 2. Fan out 3. Cross-judge 4. Pick 5. Graft 6. Verify ## Phase A: Frame The N candidates will receive the same prompt, so the prompt is the contract. Get it right before spawning anything. 1. State the artifact each candidate is producing. 2. Derive the rubric. State what success looks like for *this* task, then turn it into 3-6 concrete gradeable criteria. Concrete: `Adds a --dry-run flag that skips writes`. Vague: `code is correct`. The rubric is the picker's tool in Phase D; candidates only see the task. 3. Pick the runners. Use the parent model by default. When the host supports model selection, prefer different available model families for judgment-heavy work. Reusing the parent model is fine when generation breadth matters more than model diversity. 4. Assign output paths. Each candidate writes to its own location (a git worktree where possible, otherwise `/tmp/arena-<slug>/candidate-<n>/`). N candidates writing to the same path is shared mutable state and fails the the **separate-before-serializing-shared-state** principle skill test. ## Phase B: Fan out Use the host's native subagent tool to start all N