Critical Debater Suite

面向指定议题的多智能体辩论系统，含正方、反方、评审与编排四种角色，支持逐轮证据刷新与核验、五要素推理链与主张账本，最终产出结构化双语 Markdown 报告；运行依赖 bash、jq、python3、shasum 及网络搜索。

热度 52.34

最近更新 2026-05-11 17:31:09

分类：数据依赖：代码执行 / 读写文件 / 内部调用大模型 / 联网浏览搜索应用场景：研究 / 报告撰写 / 工作流

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add critical-debater

Metadata

name: critical-debater-suite
description: Multi-agent adversarial debate system with 4 roles (Pro, Con, Judge, Orchestrator), per-round evidence refresh, 5-element reasoning chains, and structured bilingual Markdown report output. Use this skill when the user says "debate", "start a debate", "red team this", "analyze from multiple perspectives", or provides a topic for critical examination.
license: MIT-0
compatibility: Requires bash, jq, python3, shasum, and internet access for web search.
metadata: {"author":"xwxga","version":"2.0.0","homepage":"https://github.com/xwxga/critical-debater","tags":"debate, evidence, reasoning, multi-agent, report"}

When a user request matches this skill, route to the appropriate capability module.

User Intent	Route To
"debate", "start a debate", "red team this", "analyze from multiple perspectives", full topic for debate	`capabilities/debate.md`
"search evidence", "build evidence store", "find sources"	`capabilities/source-ingest.md`
"check freshness", "update evidence status"	`capabilities/freshness-check.md`
"verify evidence", "cross-check sources"	`capabilities/evidence-verify.md`
"build arguments", "construct debate turn"	`capabilities/debate-turn.md`
"judge round", "audit round", "evaluate arguments"	`capabilities/judge-audit.md`
"update claims", "claim ledger"	`capabilities/claim-ledger-update.md`
"check analogy", "validate historical reference"	`capabilities/analogy-safeguard.md`
"final report", "synthesis", "generate report"	`capabilities/final-synthesis.md`
Mixed request	Pick primary intent, chain capabilities in required order

Use provider-agnostic names in all capability prompts:

Generic Name	Claude Code	Codex	GPT/OpenClaw
`search`	WebSearch	web_search	web_search
`fetch`	WebFetch	-	browse
`spawn_role`	Agent	subprocess	-
`read`	Read	read	read
`write`	Write	write	write

Tier	Description
`fast`	Cheapest, simple tasks
`balanced`	Good reasoning at reasonable cost
`deep`	Best reasoning, verification, causal audit

All JSON schemas are defined in references/data-contracts.md. This is the single source of truth for all agent outputs.