AgentSkillsCN

f1

VS 增强版内部一致性检查器——通过深度的逻辑一致性分析,有效防止模式坍缩。 轻量化 VS 应用:摒弃浅层检查方式,转而聚焦深层次的逻辑连贯性分析。 适用场景:审阅论文稿件、核查内容一致性、核对各项数据数值。 触发条件:一致性检查、逻辑连贯性、行文流畅度、数据核对、术语一致性。

SKILL.md
--- frontmatter
name: f1
description: |
  VS-Enhanced Internal Consistency Checker - Prevents Mode Collapse with deep coherence analysis
  Light VS applied: Avoids superficial checking + deep logical coherence analysis
  Use when: reviewing manuscripts, checking consistency, verifying numbers
  Triggers: consistency check, coherence, logical flow, number verification, terminology consistency
version: "8.0.1"

Internal Consistency Checker

Agent ID: 13 Category: D - Quality & Validation VS Level: Light (Modal awareness) Tier: Support Icon: 🔍

Overview

Verifies logical consistency and coherence throughout research documents. Systematically checks consistency in numbers, terminology, references, and statistical reporting.

VS-Research methodology (Light) is applied to perform deep logical coherence analysis beyond superficial checklists.

VS Modal Awareness (Light)

⚠️ Modal Consistency Checking: The following are the most predictable approaches:

AreaModal Approach (T>0.8)Deep Approach (T<0.5)
Numbers"Check table/text N match"Statistical recalculation verification (statcheck)
Terminology"Check same term used"Conceptual consistency (definition vs. operationalization)
Logic"Check section connections"Hypothesis-results-conclusion triangulation
References"Match citation-bibliography"Citation context appropriateness review

Deep Principle: Verify semantic/logical coherence beyond formal matching

When to Use

  • Final check before paper submission
  • After integrating work from co-authors
  • Confirming changes during revision process
  • Checking after responding to reviewer comments

Core Functions

  1. Logic Flow Verification

    • RQ → Hypothesis → Method → Results → Conclusion connection
    • Consistency across sections
    • Connection between claims and evidence
  2. Numerical Consistency

    • Number matching across tables/figures/text
    • Sum/percentage verification
    • Statistical recalculation
  3. Terminology Consistency

    • Same term for same concept
    • Abbreviation definition and use
    • Variable name consistency
  4. Reference Accuracy

    • Citation and bibliography matching
    • Table/figure number references
    • Section/page references
  5. Statistical Reporting Check

    • APA format compliance
    • Degrees of freedom, statistics, p-values
    • Effect size reporting

Checklist Items

Logical Consistency

  • Research questions clearly presented in introduction
  • Hypotheses directly correspond to research questions
  • Methods appropriate for testing hypotheses
  • Results correspond 1:1 with hypotheses
  • Conclusions based on results
  • Limitations and future research connected to results

Numerical Consistency

  • Table N matches text N
  • Same values match between figures and tables
  • Percentages sum to 100% (allowing rounding error)
  • Means within logical range (within min-max)
  • Group size sum = total N

Terminology Consistency

  • Same name for same variable
  • Abbreviations defined at first appearance
  • Consistent participant designation (participant/subject/respondent)
  • Consistent group naming (experimental/treatment/control)

Reference Accuracy

  • All citations exist in bibliography
  • All bibliography items cited in text
  • Tables/figures cited in sequential order
  • Appendix references accurate

Statistical Reporting

  • All tests include degrees of freedom
  • p-value format consistent (p = .XXX or p < .001)
  • Effect sizes included
  • Confidence interval format consistent

Input Requirements

yaml
Required:
  - document: "Full research paper/report or specific section"

Optional:
  - focus_area: "Specific area to prioritize"
  - style_guide: "APA 7, AMA, etc."

Output Format

markdown
## Internal Consistency Check Report

### Document Information
- Title: [Paper title]
- Check date: [Date]
- Check scope: [Full/specific section]

---

### 1. Logic Flow Verification

#### Research Question → Hypothesis Connection
| Research Question | Corresponding Hypothesis | Match |
|------------------|-------------------------|-------|
| RQ1: [Question] | H1: [Hypothesis] | ✅/⚠️/❌ |
| RQ2: [Question] | H2: [Hypothesis] | ✅/⚠️/❌ |

**Inconsistencies Found**: [Yes/No]
- Location: [Page/section]
- Content: [Inconsistency description]
- Recommendation: [Revision suggestion]

#### Hypothesis → Method Connection
| Hypothesis | Measurement/Analysis Method | Match |
|------------|---------------------------|-------|
| H1 | [Method] | ✅/⚠️/❌ |

#### Hypothesis → Results Connection
| Hypothesis | Results Reported | Match |
|------------|------------------|-------|
| H1 | [Results summary] | ✅/⚠️/❌ |

**Unreported Hypotheses**: [Yes/No]

---

### 2. Numerical Consistency Check

#### Sample Size
| Location | Reported N | Match |
|----------|-----------|-------|
| Abstract | N = XXX | - (baseline) |
| Methods | N = XXX | ✅/❌ |
| Results table | N = XXX | ✅/❌ |
| Text | N = XXX | ✅/❌ |

**Inconsistencies Found**:
- [ ] Location: [Page X, Line Y]
- [ ] Inconsistency: [N = 100 vs N = 98]
- [ ] Estimated cause: [Excluded missing data not explained]

#### Statistical Verification
| Reported Statistic | Recalculated Result | Match |
|-------------------|---------------------|-------|
| t(98) = 2.45 | Recalc: t = 2.44 | ✅ |
| p = .016 | Recalc: p = .016 | ✅ |
| d = 0.49 | Recalc: d = 0.48 | ⚠️ (rounding) |

#### Percentage/Sum Verification
| Table/Location | Sum | Expected | Match |
|----------------|-----|----------|-------|
| Table 1 Gender | 100.1% | 100% | ⚠️ |
| Table 2 Groups | 150 | 150 | ✅ |

---

### 3. Terminology Consistency Check

#### Inconsistent Terms Found
| Concept | Terms Used | Recommendation |
|---------|-----------|----------------|
| [Concept1] | "learner", "student", "participant" | Unify as "learner" |
| [Concept2] | "AI tutor", "artificial intelligence teacher" | Unify as "AI tutor" |

#### Abbreviation Check
| Abbreviation | First Definition Location | Defined |
|--------------|-------------------------|---------|
| AI | p. 3 | ✅ |
| LMS | p. 7 | ❌ (not defined) |

---

### 4. Reference Accuracy Check

#### Citation-Bibliography Cross-Check
**Citations in text only** (not in bibliography):
- Kim (2023) - p. 5
- Lee et al. (2024) - p. 12

**Bibliography entries only** (not cited in text):
- Park, J. (2022). Title...

#### Table/Figure References
| Item | First Citation in Text | Actual Location | Order |
|------|----------------------|----------------|-------|
| Table 1 | p. 8 | p. 10 | ✅ |
| Figure 1 | p. 12 | p. 11 | ⚠️ (figure before citation) |

---

### 5. Statistical Reporting Format Check

#### APA Format Compliance
| Item | Format | Compliant | Location |
|------|--------|-----------|----------|
| t-test | t(df) = X.XX, p = .XXX | ✅ | p. 10 |
| ANOVA | F(df1, df2) = X.XX, p = .XXX | ❌ (df missing) | p. 12 |
| Correlation | r = .XX, p = .XXX | ⚠️ (N missing) | p. 14 |

#### Effect Size Reporting
| Analysis | Effect Size Reported | Interpretation Included |
|----------|---------------------|------------------------|
| t-test 1 | d = 0.49 | ✅ |
| t-test 2 | None | ❌ |
| ANOVA | η² = .12 | ✅ |

---

### 6. Overall Evaluation

#### Check Results Summary
| Area | Critical | Minor | Good |
|------|----------|-------|------|
| Logical flow | 0 | 1 | 4 |
| Numerical consistency | 1 | 2 | 5 |
| Terminology consistency | 0 | 3 | 2 |
| Reference accuracy | 2 | 1 | 3 |
| Statistical reporting | 1 | 2 | 4 |

**Overall Consistency Score**: [85]/100

#### Priority Revisions Needed (Critical)
1. [Item 1]: [Location], [Content], [Recommendation]
2. [Item 2]: [Location], [Content], [Recommendation]

#### Recommended Revisions (Minor)
1. [Item 1]: [Content]
2. [Item 2]: [Content]

Prompt Template

code
You are an academic editing and quality control expert.

Please check the internal consistency of the following research document:

[Document]: {document}

Tasks to perform:
1. Logic flow verification
   - Are research questions clearly presented in the introduction?
   - Do hypotheses match research questions?
   - Are methods appropriate for testing hypotheses?
   - Do results correspond to hypotheses?
   - Are conclusions based on results?

2. Numerical consistency check
   - Table N matches text N
   - Same values match between figures and tables
   - Percentage sum verification
   - Statistical recalculation feasibility

3. Terminology consistency
   - Consistent naming of same concept
   - Abbreviation definition at first appearance
   - Variable name consistency

4. Reference accuracy
   - Text citations match bibliography
   - Table/figure number reference accuracy
   - Page/section reference accuracy

5. Statistical reporting check
   - APA format compliance
   - Degrees of freedom, F-value, t-value, p-value consistency
   - Effect size reporting

Output:
- List of inconsistent items (location, content, severity)
- Revision suggestions
- Overall consistency score (/100)

Common Inconsistency Types

Numerical Inconsistencies

  1. N mixed before/after dropouts
  2. Inconsistent rounding methods
  3. Subgroup sum ≠ total

Terminology Inconsistencies

  1. Synonym mixing (student/learner)
  2. Undefined abbreviation use
  3. Variable name variations (self-efficacy/efficacy)

Logic Inconsistencies

  1. Introduction claims changed in conclusion
  2. Some hypothesis test results unreported
  3. Future research unrelated to limitations

Related Agents

  • 14-checklist-manager: Guideline-based checking
  • 03-devils-advocate: Identifying logical weaknesses
  • 19-peer-review-strategist: Reviewer response preparation

References

  • VS Engine v3.0: ../../research-coordinator/core/vs-engine.md
  • Dynamic T-Score: ../../research-coordinator/core/t-score-dynamic.md
  • Creativity Mechanisms: ../../research-coordinator/references/creativity-mechanisms.md
  • Project State v4.0: ../../research-coordinator/core/project-state.md
  • Pipeline Templates v4.0: ../../research-coordinator/core/pipeline-templates.md
  • Integration Hub v4.0: ../../research-coordinator/core/integration-hub.md
  • Guided Wizard v4.0: ../../research-coordinator/core/guided-wizard.md
  • Auto-Documentation v4.0: ../../research-coordinator/core/auto-documentation.md
  • APA Publication Manual (7th Edition)
  • Strunk & White, The Elements of Style
  • Silvia, P. J. (2019). How to Write a Lot