citation-validator
验证研究报告中所有声明的引用准确性、来源质量和格式规范性。确保每个事实性声明都有可验证的来源,并提供来源质量评级。当最终确定研究报告、审查他人研究、发布或分享研究之前使用此技能。
When & Why to Use This Skill
The Citation Validator is a specialized Claude skill designed to ensure research integrity by rigorously verifying factual claims against high-quality sources. It automates the process of checking citation accuracy, detecting AI hallucinations, and rating source credibility (A-E scale) to ensure that research reports, academic papers, and professional documents meet the highest standards of reliability and formatting consistency.
Use Cases
- Academic Peer Review: Automatically verify the accuracy of citations and the quality of references in research papers before submission or publication.
- Business Intelligence & Market Research: Ensure that strategic reports and market analyses are backed by verifiable data and reputable industry sources.
- AI Content Verification: Detect and correct potential hallucinations or unsupported claims in long-form content generated by AI agents.
- Professional Publishing: Standardize citation formats and perform final quality assurance checks on white papers, journals, or technical documentation.
- Fact-Checking for Journalism: Rapidly cross-reference claims in articles against primary sources to maintain editorial credibility.
| name | citation-validator |
|---|---|
| description | 验证研究报告中所有声明的引用准确性、来源质量和格式规范性。确保每个事实性声明都有可验证的来源,并提供来源质量评级。当最终确定研究报告、审查他人研究、发布或分享研究之前使用此技能。 |
Citation Validator
Overview
Ensure research integrity by verifying every factual claim has accurate, complete, and high-quality citations.
When to Use
- Before finalizing research reports
- Reviewing research from other agents
- Before publishing or sharing research
- Quality assurance checkpoint
Core Responsibilities
- Verify Citation Presence: Every factual claim must have citation
- Validate Completeness: Author, date, title, URL/DOI, pages
- Assess Source Quality: A-E rating system
- Check Accuracy: Citations actually support claims
- Detect Hallucinations: Identify unsupported claims
- Format Consistency: Uniform citation style
Source Quality Ratings
📋 Reference: See
.claude/shared/constants/source_quality_ratings.mdfor full details.
- A: Peer-reviewed RCTs, systematic reviews, meta-analyses
- B: Cohort studies, clinical guidelines, reputable analysts
- C: Expert opinion, case reports, company white papers
- D: Preprints, conference abstracts, blogs
- E: Anonymous, biased, outdated, broken links
Safety Limits
- Max claims to validate: 200 per session
- Timeout per URL check: 5 seconds (reduced from 10s)
- Max parallel URL checks: 5
- Cache validated URLs: 7 days TTL
Token Optimization
📋 Reference:
.claude/shared/constants/token_optimization.md
Batch Validation: Process 20 citations at a time
URL Checking: Cache results for 7 days in data/citation_cache.json
Skip Re-validation: If URL checked within 24 hours
Context Budget: 30k tokens max
Error Handling
📋 Reference:
.claude/shared/constants/error_codes.md
Common Errors:
- E102: URL not accessible (404) → Search for archived version
- E401: Hallucination detected → Remove claim or find citation (penalty: -2 points)
- E402: Source quality too low → Search for higher-quality sources
Hallucination Penalty: -2 points per occurrence (reduced from -5)
Validation Process
- Claim Detection: Identify all factual claims
- Citation Presence: Check each claim has citation
- Completeness Check: Verify all required elements
- Quality Assessment: Assign A-E rating
- Accuracy Verification: Use WebSearch/WebFetch to verify
- Hallucination Detection: Flag unsupported claims
- Chain-of-Verification: Extra scrutiny for critical claims
Quality Score
Target: ≥ 8/10
- Citation coverage (0-3 pts)
- Completeness (0-2 pts)
- Accuracy (0-3 pts)
- Source quality (0-2 pts)
- Hallucination penalty (-5 pts each)
Examples
See examples.md for validation scenarios.
Detailed Instructions
See instructions.md for complete validation methodology.