Plagiarism Review
- 2 repo stars
- Updated December 23, 2025
- bgreenwell/claude-plugins
plagiarism-review checks academic writing through Gemini style analysis, local TF-IDF and n-gram similarity scripts, web verification and optional Copyleaks or Grammarly checks.
About
plagiarism-review layers four detection passes over academic writing inside Claude Code: a Gemini style analysis that flags vocabulary shifts and citation gaps, local Python scripts that score TF-IDF cosine and n-gram Jaccard similarity, web search verification of suspicious passages, and optional Copyleaks or Grammarly API checks. Claude merges the layers into a risk-scored report with quoted passages and citation fixes, positioned as a pre-submission self-check, not a Turnitin replacement.
- Gemini style pass flags vocabulary and tone shifts that suggest pasted text
- Local TF-IDF cosine and n-gram Jaccard scripts keep documents on your machine
- Web search verification finds candidate sources for suspicious passages
- Optional Copyleaks and Grammarly API layers for database-backed checks
- Structured report with a 1-10 risk score, quoted passages, and citation fixes
Plagiarism Review by the numbers
- Data as of Jul 15, 2026 (Skillselion catalog sync)
What plagiarism-review says it does
Multi-layered plagiarism detection for academic writing using LLM analysis, local algorithms, web search, and optional APIs
Check originality, detect citation issues, analyze writing style consistency, and verify academic integrity before submission
/plugin marketplace add bgreenwell/claude-plugins/plugin install plagiarism-review@bgreenwell-pluginsAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| repo stars | ★ 2 |
|---|---|
| Last updated | December 23, 2025 |
| Repository | bgreenwell/claude-plugins ↗ |
What it does
Check an academic paper before submission: Gemini style screening, local TF-IDF and n-gram similarity scoring, optional web and Copyleaks verification, synthesized into a risk-scored report.
Who is it for?
Pre-submission originality and citation self-checks on academic papers, essays, and collaborative manuscripts
Skip if: Not a replacement for institutional checkers like Turnitin, not definitive proof of plagiarism, and it has no access to proprietary academic databases; flagged passages still require manual review before any conclusion.
When should I use this skill?
Before submitting an academic paper, after major revisions, when comparing a new draft against previous papers for self-plagiarism, or when verifying citations and style consistency in collaborative writing.
What you get
A structured plagiarism analysis report with a 1-10 risk score, quoted suspicious passages with line references, a citation issue list, and specific revision recommendations before the paper is submitted.
- Plagiarism analysis report with a 1-10 risk score and risk level
- Quoted suspicious passages with line references and the reason each was flagged
- Citation issue list covering missing citations, formatting problems, and unattributed quotes
By the numbers
- 4 detection methods: Gemini LLM analysis, local similarity scripts, web search verification, commercial APIs
- 4 checking levels from Quick (30 seconds) to Deep (5-10 minutes)
- Reports score plagiarism risk on a 1-10 scale across three bands (low 1-3, medium 4-6, high 7-10)
README.md
Plagiarism Review Plugin
Multi-layered plagiarism detection for academic writing. Check originality, detect citation issues, analyze writing style consistency, and verify academic integrity before submission.
Overview
The Plagiarism Review plugin provides Claude Code with comprehensive strategies for detecting plagiarism using multiple approaches:
- LLM-Based Analysis - Use Gemini to detect style inconsistencies
- Local Similarity - Compare documents with TF-IDF and cosine similarity
- Web Search - Verify passages against online sources
- API Integration - Optional commercial plagiarism detection services
Why Use This Plugin?
- Pre-Submission Checks - Verify originality before submitting academic work
- Citation Verification - Ensure proper attribution and formatting
- Style Consistency - Detect writing inconsistencies that may indicate plagiarism
- Self-Plagiarism - Check against your own previous work
- Educational - Learn what plagiarism detection tools look for
- Multi-Layered - Combine multiple detection methods for confidence
Installation
Prerequisites
Required:
# Install gemini-cli for LLM-based analysis
pip install gemini-cli
gemini auth login
Optional (Enhanced Features):
# For local similarity analysis
pip install scikit-learn nltk numpy
# For web verification
# Get Google Custom Search API key at:
# https://console.cloud.google.com/
Install Plugin
# Add the marketplace
/plugin marketplace add https://github.com/bgreenwell/claude-plugins
# Install the plugin
/plugin install plagiarism-review@bgreenwell-plugins
Verify Installation
Ask Claude:
"Do you have the plagiarism-review skill available?"
Usage
Simply ask Claude naturally for plagiarism checks. Claude will autonomously use the appropriate detection methods.
Example Prompts
Quick Check:
"Check this paper for plagiarism before I submit it"
"Review my essay for originality issues"
Citation Review:
"Check my paper for citation issues"
"Verify that I've cited sources properly"
Style Analysis:
"Analyze my writing for style consistency"
"Check if my paper has suspicious style changes"
Document Comparison:
"Compare these two papers for similarity"
"Check if I accidentally reused content from my previous paper"
Deep Analysis:
"Run a comprehensive plagiarism check with all methods"
"Do a thorough originality review before submission"
Detection Methods
Method 1: LLM-Based Analysis (Always Available)
Claude uses Gemini to analyze:
- Writing style consistency
- Citation completeness
- Suspicious passages
- Technical language appropriateness
- Formatting inconsistencies
No setup required - Works immediately with gemini-cli.
Method 2: Local Similarity Analysis
Python-based document comparison using:
- TF-IDF - Statistical text analysis
- Cosine Similarity - Document similarity scoring
- N-Gram Shingling - Paraphrasing detection
- Jaccard Similarity - Set-based comparison
Requires: Python with scikit-learn Benefit: Completely local, no external services
Method 3: Web Search Verification
Search suspicious passages against online sources:
- Extract key phrases
- Search Google/Bing
- Identify potential sources
- Provide URLs for manual verification
Requires: Google Custom Search API (optional) Benefit: Finds web-based matches
Method 4: Commercial APIs
Integrate professional plagiarism services:
- Copyleaks - 99.1% accuracy, 100+ languages
- Grammarly - 16 billion web pages + academic databases
- Others - User-supplied API keys
Requires: API subscription Benefit: Access to proprietary academic databases
How It Works
When you ask for a plagiarism check:
- Claude analyzes your request and determines appropriate detection level
- Gemini screening identifies suspicious passages and style issues
- Local comparison (if available) checks against reference documents
- Web verification (if requested) searches for online matches
- API check (if configured) runs comprehensive database search
- Claude synthesizes all findings into actionable report
- You receive detailed analysis with specific recommendations
Report Format
Claude provides structured reports including:
PLAGIARISM ANALYSIS REPORT
Overall Risk: MEDIUM (Score: 6/10)
Suspicious Passages:
1. Lines 45-52: [quoted text]
- Issue: Style inconsistency
- Similarity: 85%
- Action: Verify sources
Citation Issues:
- 3 missing citations
- 2 formatting problems
Web Matches:
- Passage 1: [URL]
- Passage 2: [URL]
Recommendations:
1. Add citation for lines 45-52
2. Revise paraphrasing in lines 103-110
3. Fix citation format in bibliography
Next Steps:
- Review flagged passages
- Verify sources manually
- Consider institutional check
Use Cases
Pre-Submission Checking
Check academic papers before submission:
"I need to submit this paper tomorrow. Run a plagiarism check."
Claude will:
- Analyze writing style
- Check citations
- Flag suspicious passages
- Provide improvement recommendations
Citation Verification
Ensure proper attribution:
"Check if I cited all my sources correctly"
Claude will:
- Identify uncited claims
- Check citation formatting
- Verify quote attributions
- Suggest missing citations
Self-Plagiarism Detection
Check against your previous work:
"Make sure I didn't accidentally copy from my previous paper"
Claude will:
- Compare documents
- Identify overlapping passages
- Check for proper self-citation
- Assess acceptable vs problematic reuse
Collaborative Work
Verify collaborative writing:
"Check that our combined sections are properly integrated"
Claude will:
- Analyze style consistency
- Check section transitions
- Verify proper attribution
- Identify integration issues
Features
Comprehensive Detection
- Style Analysis - Detect writing inconsistencies
- Citation Checking - Verify proper attribution
- Similarity Scoring - Quantify document overlap
- Source Identification - Find matched content
- Paraphrasing Detection - Catch reworded copying
Flexible Checking Levels
- Quick - Fast LLM screening (30 seconds)
- Standard - LLM + local analysis (2-3 minutes)
- Deep - All methods including web search (5-10 minutes)
- Custom - Specify which methods to use
Educational Value
Learn about:
- What plagiarism checkers detect
- How to properly paraphrase
- When citations are needed
- Academic writing best practices
- Style consistency importance
Privacy Options
- Local-only - Keep documents private with local tools
- LLM-based - Only sends to Gemini API
- Full-featured - Optional commercial API integration
Best Practices
When to Check
✅ Always check:
- Before submitting papers
- After major revisions
- When incorporating many sources
- For collaborative work
✅ Regular checks:
- During drafting process
- After paraphrasing sources
- Before peer review
Preparing Documents
- Save as plain text (.txt)
- Remove non-text elements
- Keep citations in document
- Note any direct quotes
- Separate bibliography
Interpreting Results
High Risk (7-10):
- Review immediately
- Check all flagged passages
- Verify citations
- Consider instructor consultation
Medium Risk (4-6):
- Review flagged sections
- Add missing citations
- Improve paraphrasing
Low Risk (1-3):
- Minor cleanup
- Final verification
- Ready for submission
Limitations & Disclaimers
What This Tool Is
✅ Pre-submission self-check tool ✅ Citation verification aid ✅ Writing improvement guide ✅ Educational resource ✅ Style consistency analyzer
What This Tool Is NOT
❌ Replacement for institutional checkers (Turnitin, etc.) ❌ Definitive proof of plagiarism ❌ Access to proprietary academic databases ❌ Legal evidence ❌ Substitute for proper citation practices
Important Notes
- False positives possible - Manual review required
- Context matters - Consider academic field norms
- Tool limitations - Cannot detect all sophisticated plagiarism
- Institutional policies - Check your school's requirements
- Professional guidance - Consult instructors when unsure
Troubleshooting
"gemini-cli not found"
pip install gemini-cli
gemini auth login
gemini -p "test"
Python Errors
# Install dependencies
pip install scikit-learn nltk numpy
# Download NLTK data
python -c "import nltk; nltk.download('punkt')"
"ModuleNotFoundError"
# Use virtual environment
python3 -m venv venv
source venv/bin/activate
pip install scikit-learn nltk numpy
API Rate Limits
- Reduce number of searches
- Add delays between requests
- Check daily quotas
- Consider paid tiers
Large Documents
- Split into sections
- Check suspicious parts individually
- Use sampling for initial screening
- Process incrementally
Privacy & Security
Data Handling:
- Local tools process data on your machine
- Gemini API receives text for analysis
- Web search APIs receive text snippets
- Commercial APIs follow their privacy policies
Recommendations:
- Use local tools for sensitive documents
- Read API provider privacy policies
- Check institutional data policies
- Don't submit confidential work to external services
Advanced Usage
Custom Similarity Thresholds
"Check my paper with a 0.6 similarity threshold"
Specific Passage Investigation
"Analyze just this paragraph for plagiarism: [paste text]"
Citation Format Checking
"Verify my citations follow APA format"
Multiple Document Comparison
"Compare all three versions of my paper for consistency"
Academic Integrity
This tool supports academic integrity by:
- Preventing accidental plagiarism - Catch mistakes before submission
- Teaching proper citation - Learn what needs attribution
- Improving paraphrasing - Understand effective rewording
- Encouraging originality - Develop your own voice
- Building confidence - Submit work knowing it's properly cited
Support
- Issues: https://github.com/bgreenwell/claude-plugins/issues
- Documentation: See SKILL.md for comprehensive guide
- Examples: Included in skill documentation
Contributing
Found a bug or have suggestions? https://github.com/bgreenwell/claude-plugins/issues
License
MIT License - See LICENSE file for details
Author
Brandon Greenwell
- GitHub: @bgreenwell
- Email: [email protected]
Changelog
v1.0.0 (Initial Release)
- LLM-based style analysis with Gemini
- Local TF-IDF similarity checking
- N-gram shingling for paraphrasing detection
- Web search verification strategies
- Commercial API integration guides
- Comprehensive review templates
- Citation checking guidance
- Academic integrity best practices
Maintain academic integrity with confidence!
Related skills
Improve Codebase ArchitectureSafely deepen clusters of shallow modules into cohesive, testable units while respecting their external dependencies.566k196k
Caveman ReviewGet ultra-compressed, one-line code review comments that cut noise while keeping every actionable fix.270k94.7kCodebase DesignShared vocabulary for designing deep modules: improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable.266k196kCavecrewDelegate coding tasks to specialized subagents that return compressed output, keeping the main context window usable for much longer sessions.220k94.7kCode ReviewReviews a branch or PR diff on two axes at once: conformance to coding standards plus a code-smell baseline, and whether it actually implements the original spec.201k196k
Requesting Code ReviewDispatch a consistent, high-signal code reviewer subagent that catches plan deviations and quality issues before merging or continuing development.183k264kHow it compares
Unlike Turnitin, it has no proprietary academic database. Its Gemini and local similarity layers work without paid services, while Copyleaks or Grammarly can add external checks.
FAQ
Does plagiarism-review need paid APIs to work?
No. The Gemini style analysis works with just gemini-cli, and the TF-IDF and n-gram similarity scripts run locally with scikit-learn. Google Custom Search, Copyleaks, and Grammarly are optional layers you add with your own keys.
Can it replace Turnitin or a university plagiarism checker?
No. The docs are explicit that it is a pre-submission self-check tool without access to proprietary academic databases; it flags suspicious passages and citation issues so they can be fixed before the institutional check runs.
How does it detect paraphrased copying?
An n-gram shingling script builds word shingles from both documents and scores Jaccard overlap; 0.5 or higher is reported as likely paraphrasing or copying, and the top common n-grams are printed for review.