← 返回 Skills 市场
ljt1469

critical-thinking-framework

作者 ljt1469 · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
407
总下载
1
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install critical-thinking-framework
功能描述
基于津巴多心理学六问法,对任何新观点、主张或信息进行系统性、结构化的批判性审视。通过六个递进维度的分析,识别潜在偏差、验证逻辑有效性,并形成经过多重检验的审慎结论。适用于评估新闻报道、科学主张、政策建议及日常决策。
使用说明 (SKILL.md)

\r \r

Parameters\r

| Parameter | Type | Required | Description |\r | -------------------- | ------ | -------- | ------------------------------------------------------------ |\r | claim | string | Yes | 待分析的具体观点或主张(如"糖会导致儿童亢奋") |\r | source_info | object | No | 观点来源信息,包含author(作者)、credentials(资质)、affiliation(所属机构)、potential_interests(潜在利益) |\r | evidence_available | array | No | 当前可获取的支持性证据列表(轶事、统计数据、实验结果等) |\r | domain | string | No | 观点所属领域(医学/心理学/社会学/政策等),用于选择专业视角 |\r \r

Execution Steps\r

\r Step 1: 权威溯源与利益审查 (Source & Interest Verification)\r

  • 核查提出者是否具备相关领域的实际专业资质(非自封专家)\r
  • 识别利益冲突:是否存在政治资本、经济利益或媒体曝光动机?\r
  • 输出标记: source_credibility (high/medium/low), conflict_of_interest (boolean)\r \r Step 2: 主张合理性校准 (Claim Plausibility Calibration)\r
  • 应用"卡尔·萨根原则":不寻常的主张需要不寻常的证据\r
  • 评估与现有科学共识的兼容性,警惕"突破/革命"式宣传\r
  • 质疑"快速解决复杂问题"的承诺(如简单方法解决青少年犯罪)\r
  • 输出标记: claim_plausibility (plausible/suspicious/extreme), evidence_barrier (high/low)\r \r Step 3: 证据质量分级 (Evidence Hierarchy Evaluation)\r
  • 区分轶事证据(个人证词、精心挑选的案例)与实证证据(可重复实验、大样本研究)\r
  • 检查是否存在"将适合少数人的事物推断为适合所有人"的风险\r
  • 输出标记: evidence_type (anecdotal/correlational/experimental), generalizability_risk (high/medium/low)\r \r Step 4: 认知偏差识别 (Cognitive Bias Detection)\r
  • 情感偏差:是否存在绝望、恐惧等强烈情绪影响判断?(如绝望父母 grab any straw)\r
  • 证实性偏差:是否只记住支持信念的证据而忽略反证?(如占星术信徒的选择性记忆)\r
  • 期望偏差:观察者是否因预期而看到"想看到的结果"?(如预期糖导致亢奋而高估活跃程度)\r
  • 输出标记: detected_biases (array: [emotional_bias, confirmation_bias, expectancy_bias])\r \r Step 5: 逻辑谬误规避 (Logical Fallacy Avoidance)\r
  • 相关-因果谬误检查:是否将时间先后或统计关联等同于因果关系?\r
  • 考虑第三变量解释(C变量):是否存在同时影响A和B的隐藏因素?\r
  • 警惕"常识替代科学"(如"物以类聚"与"异性相吸"的矛盾常识)\r
  • 输出标记: causal_validity (valid/suspected_fallacy), third_variables (array)\r \r Step 6: 多元视角整合 (Multi-perspective Integration)\r
  • 强制要求至少3个不同学科视角:\r
    • 生物/心理视角:遗传、神经机制、人格特质(如糖代谢对大脑的影响)\r
    • 社会/情境视角:环境压力、群体规范、文化背景(如聚会氛围对儿童行为的影响)\r
    • 经济/系统视角:成本收益、结构性因素、激励机制\r
  • 输出标记: perspectives_considered (array), complexity_level (simple/complex)\r \r

Output Format\r

{\r
  "analysis_summary": {\r
    "overall_trustworthiness": "low/medium/high",\r
    "key_risks": ["利益冲突", "相关-因果谬误", "情感偏差"],\r
    "recommended_action": "reject/suspend_judgment/accept_with_caution"\r
  },\r
  "detailed_assessment": {\r
    "source": {"credibility": "...", "conflicts": "..."},\r
    "claim": {"plausibility": "...", "required_evidence_level": "..."},\r
    "evidence": {"quality": "...", "generalizability": "..."},\r
    "biases": ["emotional_bias", "confirmation_bias"],\r
    "logic": {"causal_validity": "...", "alternative_explanations": ["..."]},\r
    "perspectives": ["biological", "social", "economic"]\r
  }\r
}
安全使用建议
This skill is low risk: it is a descriptive checklist (no code, no installs, no credentials). Consider these points before installing: (1) provenance: the skill has no homepage and an unknown source—review outputs cautiously. (2) Input sensitivity: don't pass secrets or sensitive personal data in source_info or evidence_available because the skill will process any user-supplied content. (3) Agent actions: confirm how your agent will use the skill's outputs (e.g., avoid automatic public posting or sharing). If you need stronger assurance, run the skill in a sandboxed agent session or inspect example runs before enabling it broadly.
功能分析
Type: OpenClaw Skill Name: critical-thinking-framework Version: 1.0.0 The skill bundle is a purely analytical framework for critical thinking based on Zimbardo's psychological principles. It contains no executable code, system commands, or network requests, and its instructions in SKILL.md are limited to structured logical analysis of user-provided claims without any indicators of prompt injection or malicious intent.
能力评估
Purpose & Capability
Name/description (a six-step critical‑thinking framework) aligns with the declared parameters (claim, source_info, evidence_available, domain). No unrelated binaries, env vars, or config paths are required.
Instruction Scope
SKILL.md contains stepwise analytic instructions limited to evaluating a claim using supplied inputs. It does not instruct reading system files, environment variables, or sending data to external endpoints. Outputs are confined to structured analysis fields.
Install Mechanism
No install spec or code files — instruction‑only. Nothing is downloaded or written to disk by the skill itself.
Credentials
The skill requires no credentials, keys, or config paths. The parameters expect user-provided claim/evidence metadata only; there are no disproportionate secret requests.
Persistence & Privilege
always:false and no indication of modifying agent/system settings. The skill does not request persistent presence or elevated privileges.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install critical-thinking-framework
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /critical-thinking-framework 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release of the critical-thinking-framework skill: - Provides a structured 6-step framework, based on Zimbardo's critical thinking method, for evaluating new claims and information. - Covers source credibility, claim plausibility, evidence hierarchy, cognitive biases, logical fallacies, and multi-perspective integration. - Outputs systematic assessment results and actionable recommendations. - Supports use cases such as news evaluation, scientific claims, policy proposals, and daily decision-making.
元数据
Slug critical-thinking-framework
版本 1.0.0
许可证 MIT-0
累计安装 1
当前安装数 1
历史版本数 1
常见问题

critical-thinking-framework 是什么?

基于津巴多心理学六问法,对任何新观点、主张或信息进行系统性、结构化的批判性审视。通过六个递进维度的分析,识别潜在偏差、验证逻辑有效性,并形成经过多重检验的审慎结论。适用于评估新闻报道、科学主张、政策建议及日常决策。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 407 次。

如何安装 critical-thinking-framework?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install critical-thinking-framework」即可一键安装,无需额外配置。

critical-thinking-framework 是免费的吗?

是的,critical-thinking-framework 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

critical-thinking-framework 支持哪些平台?

critical-thinking-framework 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 critical-thinking-framework?

由 ljt1469(@ljt1469)开发并维护,当前版本 v1.0.0。

💬 留言讨论