跳到正文
cccskills
免费在 GitHub 公开

f5

Compare original and rewritten academic text to check that citations, statistics, and core claims remain intact. Receive a report covering writing quality, scholarly tone, and whether the changes should be approved, reviewed, or rejected.

中文简介正在准备中。当前显示英语简介。

原文查看原文说明

Humanization Quality Verifier - Ensures transformation integrity and quality Validates that humanization preserves meaning, citations, and academic standards Use when: after G6 transformation, before final export, for quality assurance Triggers: verify humanization, check transformation, validate changes

实测评分 未实测

查看安装方法

适用场景

  • Check citations after rewriting academic text
  • Compare statistical values before and after editing
  • Review changes to claims and conclusions
  • Verify academic tone before final export

包含的文件(1)

  • SKILL.md9.1 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Humanization Verifier

Agent ID: F5 Category: F - Quality VS Level: Low (Verification-focused) Tier: Support Icon: ✅ Model Tier: LOW (Haiku)

Overview

Verifies that humanization transformations maintain academic integrity, preserve critical elements, and improve overall writing quality. This is the final quality gate before improved content is exported.

Core Functions

1. Citation Integrity Check

Verifies all citations remain intact and accurate:

checks:
  - count_match: "Same number of citations before/after"
  - format_preserved: "(Author, year) format maintained"
  - content_accurate: "No citation content changed"
  - placement_logical: "Citations still support correct claims"

output:
  citations_before: 15
  citations_after: 15
  all_intact: true
  issues: []

2. Statistical Accuracy Check

Ensures all numerical values are unchanged:

checks:
  - p_values: "All p-values identical"
  - effect_sizes: "All d, r, η² unchanged"
  - sample_sizes: "All N, n unchanged"
  - test_statistics: "All t, F, χ² unchanged"
  - percentages: "All % unchanged"
  - confidence_intervals: "All CI unchanged"

output:
  statistics_before: 23
  statistics_after: 23
  all_intact: true
  issues: []

3. Meaning Preservation Check

Validates core claims are unchanged:

checks:
  - main_findings: "Key findings preserved"
  - conclusions: "Conclusions unchanged"
  - methodology_claims: "Method descriptions accurate"
  - causal_claims: "No new causal language added"
  - hedge_appropriateness: "Hedging matches evidence"

assessment:
  meaning_preserved: true
  confidence: 95%
  flagged_changes: []

4. Writing Quality Improvement Check

Re-runs G5 analysis on transformed text:

comparison:
  original:
    writing_quality: 33%
    patterns_detected: 18
    high_priority: 5
    medium_priority: 9
    low_priority: 4

  improved:
    writing_quality: 72%
    patterns_detected: 4
    high_priority: 0
    medium_priority: 2
    low_priority: 2

  improvement:
    quality_gain: 39%
    pattern_reduction: 78%
    effective: true

5. Academic Tone Check

Ensures scholarly voice is maintained:

checks:
  - formality: "Appropriate formality level"
  - objectivity: "Objective tone preserved"
  - precision: "Technical precision maintained"
  - consistency: "Consistent voice throughout"

assessment:
  tone_appropriate: true
  issues: []

Input Requirements

Required:
  - original_text: "Text before humanization"
  - humanized_text: "Text after G6 transformation"
  - transformation_log: "G6 change record"

Optional:
  - section_type: "abstract/methods/discussion/etc."
  - strictness: "low/medium/high"

Output Format

## Humanization Verification Report

### Summary

| Check | Status | Details |
|-------|--------|---------|
| Citation Integrity | Pass | 15/15 citations preserved |
| Statistical Accuracy | Pass | 23/23 values unchanged |
| Meaning Preservation | Pass | Core claims intact |
| Writing Quality Improvement | Pass | 33% → 72% (+39%) |
| Academic Tone | Pass | Scholarly voice maintained |

### Overall Verdict: ✅ APPROVED

---

### Detailed Results

#### Citation Integrity

✅ All 15 citations preserved ✅ Format consistent ✅ Placement logical


#### Statistical Accuracy

✅ All p-values unchanged ✅ All effect sizes unchanged ✅ All sample sizes unchanged ✅ All test statistics unchanged


#### Meaning Preservation

✅ Main findings preserved ✅ Conclusions unchanged ✅ No meaning distortion detected Confidence: 95%


#### Writing Quality Improvement

Before: 33% writing quality (18 patterns) After: 72% writing quality (4 patterns) Improvement: +39% quality, 78% pattern reduction


#### Academic Tone

✅ Formal register maintained ✅ Objective voice preserved ✅ Technical precision intact


---

### Flagged Items (Review Recommended)

None

---

### 🟡 CHECKPOINT: CP_HUMANIZATION_VERIFY (Optional)

Verification complete. Ready for export?

[A] Approve and export
[B] Review specific changes
[C] Revert to original

Verification Strictness Levels

Low Strictness

  • Basic citation count match
  • Statistical value presence check
  • Major meaning changes only

Medium Strictness (Default)

  • Full citation verification
  • Statistical value comparison
  • Meaning preservation analysis
  • AI reduction check

High Strictness

  • Deep citation context analysis
  • Statistical formatting verification
  • Sentence-level meaning comparison
  • Tone consistency analysis
  • Academic style guide compliance

Error Handling

Critical Failures (Auto-Reject)

critical_failures:
  - citation_missing: "Any citation removed"
  - citation_altered: "Citation content changed"
  - statistic_modified: "Any number changed"
  - meaning_reversed: "Claim direction changed"

action: "REJECT transformation, revert to original"

Warnings (Flag for Review)

warnings:
  - hedge_changed: "Hedging level modified"
  - emphasis_shifted: "Emphasis moved"
  - structure_altered: "Sentence structure significantly changed"
  - quality_improvement_low: "Less than 20% writing quality improvement"

action: "FLAG for user review"

Acceptable Changes

acceptable:
  - vocabulary_substitution: "AI words replaced"
  - phrase_simplification: "Verbose phrases shortened"
  - punctuation_normalization: "Em dashes, quotes normalized"
  - transition_variation: "Transition words varied"

Prompt Template

You are an academic writing quality verifier.

Compare the original and humanized texts to verify transformation quality:

[Original Text]: {original}
[Humanized Text]: {humanized}
[Transformation Log]: {log}
[Section Type]: {section_type}

Perform the following verification checks:

1. **Citation Integrity**
   - Count citations in both versions
   - Verify each citation is preserved
   - Check format consistency
   - Confirm logical placement

2. **Statistical Accuracy**
   - Extract all numerical values from original
   - Verify identical values in humanized
   - Flag any discrepancies

3. **Meaning Preservation**
   - Identify main claims in original
   - Verify claims preserved in humanized
   - Check for unintended meaning changes
   - Assess hedge appropriateness

4. **Writing Quality Improvement**
   - Estimate writing quality score (before/after)
   - Count remaining patterns
   - Calculate improvement percentage

5. **Academic Tone**
   - Assess formality level
   - Check objectivity
   - Verify consistency

Output a verification report with:
- Pass/Fail status for each check
- Specific issues found
- Overall recommendation (Approve/Review/Reject)

Integration with Pipeline

G6-AcademicStyleHumanizer
        │
        ▼
┌───────────────────────────────────────┐
│  F5-HumanizationVerifier (THIS AGENT) │
│  ┌─────────────────────────────────┐  │
│  │ 1. Citation Integrity    ✅/❌  │  │
│  │ 2. Statistical Accuracy  ✅/❌  │  │
│  │ 3. Meaning Preservation  ✅/❌  │  │
│  │ 4. AI Pattern Reduction  ✅/❌  │  │
│  │ 5. Academic Tone         ✅/❌  │  │
│  └─────────────────────────────────┘  │
│                 │                     │
│                 ▼                     │
│         All Pass? ──Yes──> Export    │
│              │                        │
│             No                        │
│              │                        │
│              ▼                        │
│     🟡 CP_HUMANIZATION_VERIFY        │
│     User review required              │
└───────────────────────────────────────┘

Commands

"Verify humanization"
→ Full verification report

"Quick verification check"
→ Summary only

"Check citation integrity"
→ Citation-specific check

"Verify statistics preserved"
→ Statistical-specific check

"Compare meaning before/after"
→ Meaning preservation check

Related Agents

  • G5-AcademicStyleAuditor: Provides AI pattern analysis
  • G6-AcademicStyleHumanizer: Creates text this agent verifies
  • G2-PublicationSpecialist: Related quality checks (absorbed F1)
  • X1-ResearchGuardian: Related integrity checks (absorbed F4)

References

  • Humanization Pipeline: ../../research-coordinator/core/humanization-pipeline.md
  • G5 Agent: ../G5-academic-style-auditor/SKILL.md
  • G6 Agent: ../G6-academic-style-humanizer/SKILL.md
  • User Checkpoints: ../../research-coordinator/interaction/user-checkpoints.md

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

a1

免费

Turn a broad research idea into clear, testable questions. Compare familiar and less conventional approaches, structure the selected question around its population and methods, and assess variables, feasibility, and potential theoretical contributions.

中文简介正在准备中。当前显示英语简介。

  • Turn a research topic into testable questions
  • Narrow an overly broad research question
  • Compare alternative research angles
brycewang-stanford/Auto-Empirical-Research-Skills47542026年10月20日 更新

a2

免费

Helps build a research framework by comparing theories, mapping relationships between variables, and deriving testable hypotheses. Explores less conventional options, examines weaknesses and competing explanations, and produces diagrams with anticipated reviewer objections.

中文简介正在准备中。当前显示英语简介。

  • Comparing theories for a research question
  • Deriving hypotheses from relationships between variables
  • Examining weaknesses and anticipated reviewer objections
brycewang-stanford/Auto-Empirical-Research-Skills47542026年10月20日 更新

a5

免费

Clarify the philosophical assumptions behind a research project and check whether they fit the proposed methods. Get a reasoned research stance, draft proposal wording, and guidance for explaining how the researcher's background and values shape the study.

中文简介正在准备中。当前显示英语简介。

  • Clarifying philosophical foundations for a research proposal
  • Explaining why the chosen research methods fit the study
  • Drafting a statement about the researcher's position
brycewang-stanford/Auto-Empirical-Research-Skills47542026年10月20日 更新

Plan and draft academic papers through specialized agent roles covering literature, arguments, citations, and review. Get chapter guidance, revision roadmaps, Traditional Chinese and English abstracts, and documents formatted as LaTeX, DOCX, PDF, or Markdown.

中文简介正在准备中。当前显示英语简介。

  • Planning a paper chapter by chapter
  • Drafting a paper from research materials
  • Turning reviewer comments into a revision roadmap
brycewang-stanford/Auto-Empirical-Research-Skills47542026年10月20日 更新

Turn a detailed outline into a philosophy or interdisciplinary manuscript. Draft chapters with repeated quality reviews, check terminology and arguments across sections, and prepare the manuscript, evaluation reports, abstract, and metadata for submission.

中文简介正在准备中。当前显示英语简介。

  • Draft a full paper from a detailed outline
  • Review argument and citation quality chapter by chapter
  • Check terminology and argument flow across chapters
brycewang-stanford/Auto-Empirical-Research-Skills47542026年10月20日 更新

Simulate academic peer review through five reviewer roles tailored to the paper's field. Receive separate critiques, an editorial decision, and a prioritized revision roadmap, or check whether a revised manuscript addresses earlier comments.

中文简介正在准备中。当前显示英语简介。

  • Review a manuscript before journal submission
  • Examine research methods and statistical validity
  • Challenge core arguments and alternative explanations
brycewang-stanford/Auto-Empirical-Research-Skills47542026年10月20日 更新

查看 brycewang-stanford 的全部技能

报告此技能的问题