Jakeschincariol/arena-skill

选用指南

AI 根据 README 整理 · 2026/10/27 04:06:33

使用方式

通过Claude Code的斜杠命令使用，如/arena，支持--agents、--quick、--seed等参数。

查看 README 依据

## Use it ``` /arena /arena --quick write the headline for our pricing page /arena --agents 32 fix the flaky test in tests/test_api.py /arena --seed 7 plan my launch week, I have 6 hours a day ``` On its own, `/arena` takes your last request as the task and the answer you did not like as the one to beat. Claude can also reach for it without the slash, when you say something like "that's a bad answer, make them compete". When it starts itself like that, it asks before it spends anything. | flag | what it does | | --- | --- | | `--agents N` | N competitors. Default 100. |

部署要求

需要Claude Code和Python 3.8或更高版本，无需安装额外依赖。

查看 README 依据

git clone https://github.com/Jakeschincariol/arena-skill.git cp -r arena-skill/skills/arena ~/.claude/skills/ ``` Or as a plugin: ``` /plugin marketplace add Jakeschincariol/arena-skill /plugin install arena-skill@arena-skill ``` Claude Code namespaces plugin skills, so installed as a plugin it shows up as `/arena-skill:arena`. Copy the folder instead if you want plain `/arena`. Project-local instead of global: copy the same folder into your repo's `.claude/skills/`. It needs Claude Code, because it spawns sub-agents with the Agent tool, and Python 3.8 or newer. Nothing to pip install.

已知限制

100个代理并非同时运行，默认并发限制为10，整个流程耗时较长；且无法修复模糊的任务描述。

查看 README 依据

against a written rubric that is in this repo for you to read and change. What you get is the strongest answer this tournament found, not a proof that it is right. That is why it shows you the attacks it survived, and why it tells you straight when the answer you rejected scored higher. **100 agents do not all run at the same moment.** Claude Code runs at most 10 tool calls at once by default (the `CLAUDE_CODE_MAX_TOOL_USE_CONCURRENCY` setting), so the arena runs in waves of 10. The spawn is 10 waves, the whole 100-agent run is 70, and it takes a while. If you raise that limit,

winning answer. Applying them is your call, and the skill asks. **The same seed gives the same cards and the same bracket.** Not the same answers. The model is not deterministic. **It does not fix a bad task.** Vague task in, 100 flavours of vague out. ## Files ``` skills/arena/SKILL.md the orchestration steps and the exact brief every sub-agent gets skills/arena/bracket.py the tournament state machine, standard library only skills/arena/strategies.json 15 reasoning modes, 12 workflows, 12 strategies. Edit freely