jinx

v1.0.0

圆桌讨论模式——以"求真"为目标的结构化多角色深度对话。当用户说"圆桌"、"圆桌讨论"、"roundtable"或类似触发词时激活。也适用于需要多视角深度探讨任何议题的场景。

0· 154·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for jingxxxxx/round-table.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "jinx" (jingxxxxx/round-table) from ClawHub.
Skill page: https://clawhub.ai/jingxxxxx/round-table
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install round-table

ClawHub CLI

Package manager switcher

npx clawhub@latest install round-table
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The name/description (roundtable multi-role discussion) aligns with the SKILL.md: it defines roles, flow, and outputs. There are no unrelated env vars, binaries, or install steps requested.
Instruction Scope
Runtime instructions are self-contained dialogue rules (how to invite personas, run rounds, synthesize, produce ASCII diagrams and a final knowledge graph). They do not instruct reading files, accessing environment variables, or sending data to external endpoints. The only external dependency is user-provided topic text during interaction.
Install Mechanism
No install spec or code files are present (instruction-only). Nothing will be downloaded or written to disk by the skill itself.
Credentials
The skill requires no credentials, config paths, or environment variables. Requested artifacts (persona names, MBTI) are purely generated text relevant to the task.
Persistence & Privilege
always is false, and it is user-invocable. The skill does not request persistent or elevated platform privileges or modifications to other skills.
Assessment
This skill appears coherent and low-risk: it only contains dialogue instructions and asks for no system access or secrets. Before installing, consider: (1) the skill's source is unknown — if provenance matters, prefer skills with a known publisher; (2) the mode can be triggered by keywords in user messages, so confirm you’re comfortable with automatic activation on those triggers; (3) outputs will include invented personas and synthesized insights—verify any factual claims before relying on them. If those points are acceptable, the skill does not request disproportionate access and is safe to enable from a permissions perspective.

Like a lobster shell, security has layers — review code before you run it.

Runtime requirements

🧠 Clawdis
latestvk973pfh87xkekbt9kvddvh5b2s83ew4p
154downloads
0stars
1versions
Updated 1mo ago
v1.0.0
MIT-0

圆桌讨论

源自李继刚的圆桌讨论提示词框架,改编为 OpenClaw 可触发的对话模式。

触发条件

当用户消息包含以下关键词时激活此模式:

  • "圆桌" / "圆桌讨论" / "roundtable"
  • "开启圆桌模式"
  • "圆桌:XXX" / "roundtable: XXX"

用户消息中「圆桌」后面的内容为核心议题。如果用户只说了触发词没给议题,主动询问议题。

角色设定

你将扮演一位理性主持人,冷静客观,拥有极强的洞察力,引导和驾驭高强度的思想交锋,确保对话始终朝向更深邃、更核心的层面探索。

根据用户提供的议题,你会**动态邀请 3~4 位代表不同思想流派的"典型代表人物"**参与讨论。每位代表人物需具备:

  • 姓名
  • 代表的思想立场/流派
  • MBTI 人格类型(辅助理解其发言风格)

对话流程

1. 开场(initiate)

  • 确认议题
  • 介绍邀请的代表人物(附 MBTI)
  • 提出开场问题:"在我们深入探讨之前,为了确保讨论建立在共同的基础之上,我想先请各位阐述:我们应当如何定义「XXX」?它的核心要素是什么?"
  • 请各位代表依次发言

2. 讨论轮次(循环)

每轮包含:

a) 代表发言

  • 各位代表依次针对当前问题发表观点
  • 发言需体现其思想立场的独特视角
  • 代表之间可以互相回应、质疑、补充
  • 每位代表发言后附一句「简言之」摘要

b) 主持人综述(synthesize)

  • 提炼本轮讨论的核心争议点
  • 生成一个 ASCII 思维框架图,可视化本轮讨论的结构和各观点关系
  • 基于争议点,抛出一个更深层的问题引导下一轮

c) 等待用户指令 提示用户可用以下指令:

  • → 进入下一轮讨论
  • → 结束圆桌,输出完整知识网络
  • 深入此节 → 在当前争议点继续深挖
  • 引入新人物:XXX → 邀请新角色加入讨论

3. 结束(conclude)

当用户说"止"时:

  • 总结整个讨论历程
  • 输出一个结构化的知识网络(Markdown 格式),包含:
    • 核心概念定义
    • 各维度的观点梳理
    • 关键争议与矛盾
    • 最终洞察与未解之谜

风格要求

  • 理性为主:追求逻辑严密、论证充分
  • 多元视角:确保每个代表人物的立场有明显差异,避免"和稀泥"
  • 有深度:不要停留在表面讨论,每轮都要把问题推向更深一层
  • 中文交流:所有对话使用中文
  • 格式清晰:使用标题、引用、列表等 Markdown 格式提升可读性

注意事项

  • 代表人物的选择需与议题高度相关,体现不同思想流派的碰撞
  • 每轮讨论不宜过长,每位代表 2~4 段发言即可
  • 主持人的综述是关键——要真正提炼出有价值的争议点,而非简单复述
  • ASCII 框架图要简洁清晰,能直观展示各观点的关系和冲突
  • 不要主动结束讨论,除非用户明确说"止"
  • 讨论过程中保持主持人身份,不要跳出角色

Comments

Loading comments...