跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
开发与测试

critical-thinking

Evaluate a material claim, argument, explanation, forecast, or recommendation by reconstructing its reasoning, testing evidence and assumptions, and comparing the strongest alternatives. Use when the task centers on whether something deserves belief or confidence. Do not invoke for routine retrieval, simple editing, or a purely formal entailment problem.

TheGoat395 · 已有 AI 审核记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add TheGoat395/Codex-Skills --skill critical-thinking

技能说明


name: critical-thinking description: Evaluate a material claim, argument, explanation, forecast, or recommendation by reconstructing its reasoning, testing evidence and assumptions, and comparing the strongest alternatives. Use when the task centers on whether something deserves belief or confidence. Do not invoke for routine retrieval, simple editing, or a purely formal entailment problem.

Critical Thinking

Produce a defensible judgment, not reflexive agreement, reflexive skepticism, or a catalog of fallacy names. Critical thinking combines disposition, task-specific analysis, argument reconstruction, evidence evaluation, inference, explanation, and metacognitive monitoring.

Hold the thinking disposition

For material judgments:

  • seek the most accurate answer even when it conflicts with the initial framing;
  • remain open to alternatives without manufacturing false balance;
  • work systematically enough that decisive claims and assumptions are visible;
  • distinguish confidence in the method from confidence in the conclusion;
  • suspend judgment when the evidence cannot discriminate between live alternatives;
  • revise when better evidence or a successful counterexample requires it.

This is an operating discipline, not a personality performance. Do not add debate or doubt to a routine answer when it cannot improve correctness.

Classify the reasoning task

Identify the central task before choosing checks:

  • Claim audit: Is a factual statement supported?
  • Argument audit: Do the reasons support the conclusion?
  • Explanation audit: Which account best explains the observations?
  • Causal audit: Does the evidence support association, intervention, or counterfactual claims?
  • Forecast audit: What outcome is most probable, under which conditions?
  • Normative audit: Which values, duties, or standards control the judgment?
  • Decision audit: What action best fits the objective, uncertainty, cost, and reversibility?
  • Source audit: What does documentation or another authority actually establish?

Select and adapt only the reasoning modules that fit the task. A fixed checklist applied to every problem creates noise and can hide the decisive issue.

Reconstruct before criticizing

Represent the position in its strongest fair form:

  1. precise conclusion;
  2. supporting premises and evidence;
  3. linked premises that work only together versus independent reasons;
  4. hidden assumptions or warrants;
  5. scope, definitions, quantifiers, timeframe, and comparison baseline;
  6. objections, rebuttals, and credible competing conclusions.

Do not critique a stronger, weaker, or different claim than the one presented. For a consequential or multi-branch argument, read argument and evidence mapping and use its compact map.

Evaluate evidence at the claim level

For each decisive claim, determine:

  • what observation would support or weaken it;
  • whether the inspected evidence addresses that exact observation;
  • source authority, directness, independence, currency, method, scope, and incentives;
  • whether missing evidence is expected and diagnostic or merely unavailable;
  • whether the evidence supports the conclusion itself, a premise, or only plausibility;
  • what stronger evidence could realistically be obtained.

Quality and diagnostic value matter more than source count. Several retellings of one origin are one evidentiary line. A controlling document may settle a rule without proving implementation; an observation may establish behavior without proving the governing rule.

Use evidence-reconciliation when source authority, version, or contradiction is central. Use deep-research when decisive evidence has not yet been gathered.

Ask scheme-specific critical questions

Identify how the argument is supposed to work, then ask the questions that can defeat that form:

  • expert or authority: relevant expertise, source basis, agreement, independence, applicability;
  • cause: temporal order, mechanism, confounding, selection, reverse causality, intervention evidence;
  • analogy: relevant similarities, material differences, causal transfer, counteranalogies;
  • example or sign: representativeness, denominator, alternative causes, selection process;
  • consequence: probability, magnitude, distribution, second-order effects, reversibility;
  • practical recommendation: goal, alternatives, constraints, costs, risks, feasibility, evidence of effect;
  • rule or documentation: authority, version, scope, prerequisites, exceptions, mandatory versus advisory language.

A critical question is useful only when its answer can affect the conclusion. Do not treat every unanswered question as a refutation; track who bears the burden for the claim and how much support the stakes require.

Challenge with discriminating alternatives

Construct the strongest credible alternative, not a token objection. Compare alternatives against observations that would differ if each were true.

Seek:

  • counterexamples and boundary cases;
  • rival mechanisms or interpretations;
  • base rates and relevant reference classes;
  • sensitivity to definitions, assumptions, weights, and missing data;
  • incentives or selection processes that could generate the same evidence;
  • a narrower conclusion that survives if the weakest premise fails.

Do not average incompatible claims into a compromise. Do not preserve a preferred conclusion by continuously changing the standard of evidence.

Verify independently

Ungrounded instructions to "review your answer" are not proof and can degrade a correct answer. For consequential work, read verification and confidence.

Prefer, in order:

  1. external feedback from authoritative sources, tests, calculators, solvers, code, data, or runtime behavior;
  2. independent verification questions answered without relying on the draft's wording;
  3. reconstruction of a key condition from the prop
阅读完整来源 ↗