convez.ai
SkillEscrita & ConteúdoIniciante
Lucasli6833

Slop Cop - Detector de Texto Gerado por IA

Lucasli6833/slop-cop

Detecta texto gerado por IA em documentos para manter qualidade humana

Ver no GitHub
InicialPluginScore · de 0 a 100

Ferramenta que identifica padrões de linguagem gerada por IA em seus documentos, analisando vocabulário, estrutura de sentenças e assinaturas de modelos específicos. Ajuda a garantir que seu conteúdo soe autêntico e genuinamente humano antes de publicar.

Instalação

$npx convez-ai Lucasli6833/slop-cop

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
4
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 1 mês
Versão
—
Issues abertas
0
Contribuidores
3

Casos de uso

  • Verificar autenticidade de artigos e conteúdo antes de publicação
  • Revisar textos de equipes para identificar uso excessivo de geradores de IA
  • Melhorar qualidade de escrita detectando padrões robóticos

Veredito

Ferramenta prática para quem precisa garantir autenticidade em conteúdo escrito, com interface acessível. Limitação: funciona apenas em Windows e requer análise local, sem integração nativa com fluxos de trabalho em nuvem.

README

A tradução para o português deste README ainda está na fila.

🛡️ slop-cop - Identify AI writing before you publish

Download slop-cop

📁 What is this tool

Slop-cop finds AI-generated text in your documents. If you worry that your work sounds like a robot wrote it, this tool helps. It scans your writing for patterns often found in text from systems like ChatGPT, Claude, and Gemini. It looks at the specific words, the structure of your sentences, and how you organize your ideas.

The software uses data from scientific papers to learn how humans write naturally. It compares your text against these patterns to give you a clear report on whether your writing sounds authentic or machine-made.

⚙️ How it works

The tool operates on three main levels:

  1. Vocabulary analysis: It flags words that AI models overuse.
  2. Structure patterns: It identifies sentence arrangements that suggest machine generation.
  3. Model fingerprinting: It recognizes specific writing styles tied to known AI providers.

You do not need to understand how the code works. The tool handles the analysis in the background. It provides simple feedback so you can revise your drafts until they sound like a person wrote them.

📥 Getting the software

You need to access the release page to install the application on your computer.

Click here to visit the download page for slop-cop

💻 System requirements

Your computer must meet these basic requirements to run the software correctly:

  • Operating System: Windows 10 or Windows 11.
  • Memory: At least 4 gigabytes of RAM.
  • Storage: 200 megabytes of free space on your hard drive.
  • Internet Connection: Required for the first run to ensure all patterns update.

🚀 Installation guide

  1. Visit the download page.
  2. Look for the section labeled "Releases" on the right side of the screen.
  3. Click the link for the latest version.
  4. Download the file named "slop-cop-setup.exe".
  5. Locate the file in your Downloads folder.
  6. Double-click the file to start the installation.
  7. Follow the prompts on your screen. Click "Next" to continue.
  8. Accept the default location for the installation.
  9. Click "Install" to finish the process.

📝 Running an analysis

Once you install the tool, you can start checking your files immediately.

  1. Open the "slop-cop" shortcut from your start menu or desktop.
  2. Click the "Browse" button in the application window.
  3. Select the document you wish to check. The tool accepts text and common document formats.
  4. Click the "Scan" button.
  5. Wait for the progress bar to finish.
  6. Review the results window.

The tool highlights sections of text that look like AI output. It explains why it flagged those sections. You can then choose to edit those parts or adjust your writing style to improve the human score of your document.

💡 Best practices for clear writing

While slop-cop helps flag machine patterns, these habits lead to better writing:

  • Use active voice: Say "The team finished the report" instead of "The report was finished by the team."
  • Vary sentence length: Mix short, punchy sentences with longer, descriptive ones.
  • Add personal experience: Include specific anecdotes that a model would not know.
  • Avoid generic transitions: AI tends to use transition words like "furthermore" or "moreover" in excess. Use direct links between ideas instead.

🛠️ Troubleshooting common issues

If you encounter a problem while using the tool, try these steps first:

  • Application does not open: Ensure that you installed all Windows updates.
  • Scan fails: Verify that your file is not open in another program at the same time.
  • Slow performance: Close other browser tabs or heavy programs while you scan large documents.

If the issues persist, restart your computer and try opening the application again. The tool creates a small log file in your documents folder if an error occurs. You can look at this file to see if a specific file type is causing the conflict.

🚩 Understanding the results

The report screen displays a percentage score.

  • High human score: The text contains complex sentence structures and varied vocabulary.
  • Low human score: The text follows rigid, repetitive patterns.

The tool provides a list of specific "tells" found in your text. A "tell" is a phrase or a specific way of linking sentences that frequent the models we studied. If you see many tells in one paragraph, rewrite that section using your own voice.

🔒 Privacy and security

The tool performs all analysis on your local machine. It does not send your documents to a company server. Your data stays on your hard drive. This ensures that your private drafts stay private. You can run the program even when you have no internet connection, as it stores all the detection data locally during the installation.

Ver o original no GitHub

Bom para

Relacionados