TheGoat395/Codex-Skills / skills/critical-thinking
critical-thinking
VerifiedJudge claims through evidence and alternatives.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 4.9 KB |
| agents/openai.yaml | config | 295 B |
| references/argument-evidence-mapping.md | markdown | 5.2 KB |
| references/argument-review-protocol.md | markdown | 1.8 KB |
| references/challenge-protocol.md | markdown | 1.6 KB |
| references/verification-confidence.md | markdown | 4.3 KB |
references/argument-evidence-mapping.md
View on GitHub# Argument And Evidence Mapping
Use this reference when a claim depends on multiple premises, objections, sources, causal links, or competing explanations. The goal is to expose the structure that prose can hide.
## Compact argument map
Represent the argument as:
```text
Question:
Conclusion C:
R1 supports C because ...
P1 evidence/status:
P2 evidence/status:
W1 hidden warrant:
R2 independently supports C because ...
O1 objects to C because ...
Reply or unresolved effect:
Alternative conclusion A:
Discriminating observation:
```
Use atomic propositions. One node should make one claim. Split a sentence when different parts have different evidence or confidence.
## Map support correctly
- **Linked premises:** P1 and P2 support C only together. If one fails, the linked reason fails.
- **Convergent reasons:** R1 and R2 independently support C. Failure of one does not erase the other.
- **Serial inference:** P supports I, and I supports C. Test both steps.
- **Objection:** evidence against C, a premise, a warrant, or applicability.
- **Rebuttal:** a response to a specific objection, not another unrelated reason for C.
Do not count linked premises as multiple independent lines of evidence. Do not count two sources that repeat the same origin as convergent support.
## Classify the claim
The critical questions depend on claim type:
| Claim type | Main burden | Frequent failure |
|---|---|---|
| Descriptive | Direct, representative observation | Scope drift or selected examples |
| Comparative | Same definitions and conditions | Unequal baselines or omitted denominator |
| Causal | Identification of intervention effect | Correlation, confounding, reverse causality |
| Forecast | Calibrated model and reference class | Base-rate neglect or false precision |
| Explanatory | Better fit than live alternatives | Story coherence mistaken for evidence |
| Normative | Explicit values, duties, or standards | Hidden value premise or category error |
| Practical | Goal fit, alternatives, consequences | Treating one feasible action as necessary |
| Documentary | Controlling authority and applicability | Quoting text outside version or scope |
## Evaluate each evidence edge
For each source-to-claim edge, record:
| Field | Question |
|---|---|
| Relevance | Does this evidence bear on this exact claim? |
| Directness | Observation, primary record, analysis, or retelling? |
| Authority | Does it control the rule or merely discuss it? |
| Method | How were data selected, measured, and analyzed? |
| Independence | Does it share an origin or incentive with other evidence? |
| Scope | Population, environment, timeframe, version, and exceptions? |
| Diagnosticity | Would the evidence be substantially different if an alternative were true? |
| Status | Verified, supported inference, disputed, conditional, or unknown? |
Keep unknown separate from negative evidence. Absence becomes evidence only when the observation was likely to exist, be detectable, and be preserved if the claim were true.
## Apply argument-scheme questions
### Expert opinion
- Is the source an expert in the exact domain?
- What evidence or method underlies the opinion?
- Is the opinion represented accurately and in context?
- Do relevant independent experts disagree, and why?
- Does an incentive, conflict, or authority limitation matter?
### Analogy
- Which shared properties perform the inferential work?
- Are those properties causally or structurally relevant?
- Which differences could break transfer?
- Is there a closer counteranalogy?
### Cause
- Is the proposed cause prior to the outcome?
- What mechanism links them?
- What common causes, selection effects, mediators, or reverse paths exist?
- What intervention or natural experiment would discriminate the claim?
- At what causal level is the evidence: association, intervention, or counterfactual?
### Consequences and practical reasoning
- What exact goal or value controls the recommendation?
- What alternatives, including status quo and staged options, exist?
- How likely and material are benefits, harms, and second-order effects?
- Who receives each outcome?
- What constraints, prerequisites, and reversibility affect the choice?
## Burden and sufficiency
The claimant normally bears the burden for a positive assertion, but burdens can shift when a credible objection supplies evidence. Match the required support to consequence, reversibility, prior plausibility, and available evidence.
Do not demand impossible certainty for an ordinary reversible choice. Do not accept mere plausibility for a high-impact irreversible claim.
## Research basis
- Peter Facione, *Critical Thinking: A Statement of Expert Consensus* (1990): interpretation, analysis, evaluation, inference, explanation, and self-regulation.
- Diane Halpern, *Teaching Critical Thinking for Transfer Across Domains* (1998): disposition, explicit skills, structural transfer, and metacognitive monitoring.
- Christopher Dwyer, Michael Hogan, and Ian Stewart, *An Evaluation of Argument Mapping* (2012): argument-mapping instruction improved measured critical-thinking performance in the reported study.
- Douglas Walton and David Godden, *The Nature and Status of Critical Questions in Argumentation Schemes*: scheme-specific questions and defeasible burdens.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |