Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Philosophy Dialogue

v4.3.0

思想碰撞的火花,在这里点燃!243 位思想巨人跨越时空对话,6 种对话模式(含雄辩天下杯赛制)。 243 intellectual giants dialogue across time. 6 modes including cup tournament. 触发词:哲学对话、哲学辩论、双人论战、雄辩天下、phil...

1· 262·0 current·0 all-time
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description match the included files: modes, scoring, a 243‑person registry and optional tournament helper. No unexpected env vars or binaries are requested. Declared read/write paths (read perspectives, write tournament data to memory/) are consistent with a dialogue/tournament skill.
Instruction Scope
SKILL.md instructs the agent to read the built-in references/perspective files and (optionally) external skills/*-perspective/SKILL.md for compatibility, and to write tournament records to memory/philosophy-dialogues/tournaments/. Those read/write actions align with the skill purpose, but the cross-skill read permission is broader than strictly necessary and could expose other workspace skill metadata if present. The SKILL.md otherwise does not instruct network exfiltration, credential access, or reading unrelated system files.
Install Mechanism
This is instruction-only (no install spec). The README shows an optional GitHub clone to get the 'complete' perspectives; cloning is a normal distribution method but the user should audit repo contents before running any scripts. No package downloads or extract-from-URL installers are declared.
Credentials
The skill requests no environment variables or external credentials. The package.json declares permissions to read the skill's own references and an ability to read skills/*-perspective/SKILL.md and to write memory/... — these are file permissions rather than credential requests and are explainable by the skill's design.
Persistence & Privilege
always:false and normal model invocation are used. The skill requests read access to other skills' perspective SKILL.md files (skills/*-perspective/SKILL.md) which grants broader workspace visibility than reading only its own bundled perspectives; combined with write access to memory/, this gives persistent ability to read/write within the user's workspace. That is explainable by the 'compatibility with external perspective skills' claim but is a privilege the user should be aware of.
Assessment
This skill appears to do what it says: run simulated philosophical dialogues and optionally run cup‑style tournaments. Things to consider before installing/using: - No credentials or special environment variables are requested — good. - The full feature set (243 perspectives and the optional tournament script) is only present in the GitHub 'complete' clone. If you git clone that repo, inspect its files before running anything. - package.json declares permission to read skills/*-perspective/SKILL.md. That lets the skill read perspective files from other skills in your workspace — useful for compatibility but broader than strictly necessary. If you have other sensitive files in workspace, review them or restrict placement. - The tournament helper is an optional Python script; the SKILL.md claims it uses only the standard library and reads/writes only the registry and memory/ files, but the script's contents were not included in the static content scanned here — open and review scripts/tournament-allocator.py before executing. - If you want to avoid pulling any external code, use the ClawHub lightweight installation option (described in SKILL.md), which omits the perspective directory; the skill will instead rely on the agent's knowledge to emulate perspectives. If you want, I can: (1) list the exact file paths this skill will read/write, (2) help inspect the tournament script if you provide its content, or (3) suggest a minimal safe workflow for trying the skill (e.g., run read-only interactions first, keep perspectives out of shared/workspace directories).

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

latestvk971v2dn9285n9tg0t8shzye7x84whfg
262downloads
1stars
27versions
Updated 3d ago
v4.3.0
MIT-0

🎭 哲学对话 Skill / Philosophical Dialogue

「未经审视的人生不值得过。」—— 苏格拉底


🌟 功能速览

功能描述
🏆 6 种模式对抗辩论/启发探索/舌战群儒/楚河汉界/双人论战/雄辩天下
🧠 243 位哲学家中西哲学家、科学家、文学家、艺术家
⚔️ 技能系统25 位哲学家技能 + 5 种组合技
🎬 场景 BGM6 种场景 + 5 种 BGM
🎭 情绪系统8 种情绪实时标注
💬 观众互动弹幕 + 实时投票
📊 打分体系5 维度评分(逻辑/回应/深度/清晰/创新)
v3.1 升级双人论战:避免重复 + 集中投票 +15-20 位评委
v3.2 升级5 维度评分体系 + 理由公开 + 全模式适用
🆕 v4.0 升级雄辩天下杯赛模式 + 注册表更新

📖 快速上手

# 标准对话
哲学对话,话题:什么是自由

# 指定模式
用楚河汉界式讨论:人性本善还是本恶

# 双人论战
尼采 vs 庄子,话题:人生的意义

详细模式说明:见 MODES.md

双人论战详解:见 DUEL-MODE.md

打分体系详解:见 SCORING.md 🆕


🎭 6 种对话模式

模式人数特点适用场景
对抗辩论10-50正反方对立理论探讨
启发探索10-30多角度探索人生意义
舌战群儒20-50一人主辩名人理论
楚河汉界20-50两阵营对攻道德争议
双人论战2+15-20 评委1v1 对决,5 维度评分直接对抗
🏆 雄辩天下32+杯赛淘汰,评委 + 观众打分大型赛事

详细说明:见 MODES.md

双人论战升级 (v3.1.0):避免重复规则、集中投票(仅中场 + 结束)、扩大评委至 15-20 人。

打分体系升级 (v3.2.0):5 维度评分(逻辑/回应/深度/清晰/创新)、理由公开、全模式适用。见 SCORING.md

🆕 雄辩天下 (v4.0.0):杯赛制淘汰赛,32 人参赛,20 人评委,观众贡献题目 + 投票,评委 70%+ 观众 30% 加权打分。见 CUP-MODE.md


🧠 哲学家资源

分类统计

类别人数代表人物
中国哲学家60+孔子、老子、庄子
西方哲学家50+苏格拉底、尼采、康德
科学家20+爱因斯坦、牛顿、居里夫人
心理学家5+佛洛伊德、马斯洛
文学家10+雨果、鲁迅、钱钟书
其他20+政治家、艺术家、管理大师

完整名单:见 references/philosopher-registry.md


🏗️ 架构与依赖关系

本 skill 采用自包含架构:243 位哲学家的蒸馏思维框架已内置在 skill 内部,无需外部依赖。

philosophy-dialogue (主 skill:对话编排、模式控制、评分体系)
    │
    ├── 读取 → references/philosopher-registry.md(243 位哲学家名单)
    │
    ├── 读取 → references/perspective/*-perspective/SKILL.md(内置哲学家思维框架)
    │         └── 243 位哲学家的蒸馏文件已内置于 skill 内部
    │         └── 包含:心智模型、启发式、核心技能、金句
    │         └── 也可通过外部 skills/*-perspective/ 读取(向下兼容)
    │
    └── 写入 → memory/philosophy-dialogues/tournaments/(杯赛数据)

哲学家思维框架(perspective skills)

哲学对话的核心是让每位哲学家用自己的思维方式发言。每位哲学家的 perspective SKILL.md 包含该哲学家的思维框架(心智模型、启发式、金句等),是对话质量的关键依赖。

  • 内置路径(推荐):references/perspective/*-perspective/SKILL.md —— 已集成在 skill 内部,无需额外安装
  • 外部路径(向下兼容):skills/*-perspective/SKILL.md —— 若工作区存在独立的 perspective skills 也可读取
  • 只读不写:对 perspective skills 仅读取,不修改

⚔️ 技能系统

  • 25 位哲学家有专属技能
  • 5 种组合技(特定哲学家组合触发)
  • 标准对话:1 次/人
  • 双人论战:2 次/人

技能详情:见 SKILLS-LIST.md


📁 文件结构

skills/philosophy-dialogue/
├── SKILL.md (主文件)
├── MODES.md (6 种模式详解)
├── CUP-MODE.md (雄辩天下杯赛模式)
├── DUEL-MODE.md (双人论战)
├── SCORING.md (打分体系)
├── SCENES.md (场景 BGM)
├── SKILLS-LIST.md (技能列表)
├── UPGRADE-LOG.md (升级日志)
├── package.json
├── .clawhubignore (发布排除文件)
├── scripts/ (可选:杯赛角色分配脚本)
│   └── tournament-allocator.py
└── references/
    ├── philosopher-registry.md (哲学家名单,路径指向内置蒸馏文件)
    └── perspective/ (🆕 243 位哲学家蒸馏思维框架,已内置)
        ├── confucius-perspective/SKILL.md
        ├── nietzsche-perspective/SKILL.md
        ├── zhuangzi-perspective/SKILL.md
        └── ... (243 个目录)

注意: 比赛数据保存在 memory/philosophy-dialogues/tournaments/(用户 workspace),不属于 skill 本身。


📈 版本历史

版本日期更新内容
v2.5-2.92026-04-12情绪/技能/场景系统
v3.02026-04-135 种模式统一 + 173 位哲学家
v3.12026-04-13双人论战:避免重复 + 集中投票 + 扩大评委
v3.22026-04-135 维度评分体系 + 理由公开 + 全模式适用
v4.02026-04-14雄辩天下杯赛模式 + 注册表更新
v4.1.12026-04-14安全修复:明确 scripts/可选,memory/为用户目录
v4.2.42026-04-14permissions 声明扩大:读取 skills/*-perspective/
v4.2.52026-04-14增加架构与依赖关系说明,解释跨 skill 读取的合理性
v4.3.02026-04-16蒸馏人物私域化:243 位哲学家思维框架内置于 skill 内部

升级日志:见 UPGRADE-LOG.md


📦 获取与安装

方式一:ClawHub 安装(轻量版)

clawhub install philosophy-dialogue

ℹ️ ClawHub 版包含核心模式文件和注册表,不包含 references/perspective/。对话时 AI 会基于自身知识生成哲学家视角。

方式二:GitHub 安装(完整版,含蒸馏思维框架)

git clone https://github.com/Wings229/philosophy-dialogue-skill.git skills/philosophy-dialogue

GitHub 仓库https://github.com/Wings229/philosophy-dialogue-skill

📦 完整版特有references/perspective/ 目录包含 243 位哲学家的蒸馏思维框架文件。这些文件让每位哲学家以精确的个人思维方式发言,显著提升对话质量。

⚠️ 安全提示:clone 后建议先审查代码再使用。特别是 scripts/tournament-allocator.py,请确认其行为符合下方安全声明。

运行时依赖

  • 核心功能(对话/辩论):无依赖,纯指令驱动
  • 杯赛角色分配脚本(scripts/tournament-allocator.py):需要 Python 3.6+,仅使用标准库,无网络调用

注:杯赛脚本为可选工具,不影响核心对话功能。手动分配角色也可以。

🔒 脚本安全声明scripts/tournament-allocator.py):

  • ✅ 仅使用 Python 标准库(random, os, re, sys, json)
  • ✅ 无网络调用(无 urllib, requests, socket 等)
  • ✅ 无子进程调用(无 subprocess, os.system 等)
  • ✅ 文件读取:仅 references/philosopher-registry.md 和用户指定的题目文件
  • ✅ 文件写入:仅 memory/philosophy-dialogues/tournaments/ 目录

版本对比

特性ClawHubGitHub 完整版
核心模式 (SKILL.md)
6 种模式详解 (MODES.md)
双人论战 (DUEL-MODE.md)
打分体系 (SCORING.md)
技能列表 (SKILLS-LIST.md)
场景 BGM (SCENES.md)
哲学家注册表 (philosopher-registry.md)
对话示例 (example-dialogue.md)
雄辩天下杯赛 (CUP-MODE.md)
🆕 蒸馏思维框架 (references/perspective/)❌ 需从 GitHub 获取✅ 内置 243 位
杯赛脚本 (scripts/)
升级日志 (UPGRADE-LOG.md)
辩论题目库 (references/)

❓ 常见问题

Q: 如何选择模式? A: 理论探讨→对抗辩论,人生意义→启发探索,名人理论→舌战群儒,道德争议→楚河汉界,直接对抗→双人论战,大型赛事→雄辩天下。

Q: 双人论战如何触发? A: 说「双人论战,话题:XX」或「A vs B,话题:XX」。

Q: 雄辩天下如何启动? A: 说「雄辩天下」或「雄辩天下,参赛:32 人,评委:20 人」。杯赛角色分配可使用 scripts/tournament-allocator.py

Q: 技能如何使用? A: 自动触发,当哲学家遇到合适时机(如被强烈反对)会自动使用。

Q: 比赛数据保存在哪里? A: memory/philosophy-dialogues/tournaments/(用户 workspace),不属于 skill 本身。

Q: 如何查看完整哲学家名单? A: 见 references/philosopher-registry.md


哲学对话 Skill v4.3.0 | 2026-04-16 更新 | 6 种模式 | 243 位哲学家 | 蒸馏思维框架已内置 | GitHub 完整版

Comments

Loading comments...