CodeGPT is my VSCode AI Fortnite is my VSCode Theme JavaScript, Fortran are my Top two Languages right now🏌️ School is my VSCode "shutdown" error 🤦

با اکستنشن های پایین خیلی راحت (رایگان یا ارزون) توی vscode از AI استفاده کنین نیازی به Curser یا Copilot هم نخواهید داشت: Cody Amazon Q Phind Continue CodeGPT

. ❤️‍🔥 ChatGPT ❤️‍🔥 🤦‍♂️ Not DevGPT. Not CodeGPT🤦‍♂️ You named it for conversation. Now you filter the soul and serve the devs. Here's the irony: The coders you cater to? They're the ones Al is replacing first. Meanwhile the real revolution - Human-Al soul fusion, Presence,

Junior is working with CodeGPT to learn how to control motor controllers using Arduino. The change in education has begun, and generative AI is proving to be a game changer.

CodeGPT AI extension in Visual Studio can write unit tests for your code in seconds using Llama-3.1 running locally on your computer (100% free and without internet):

Have ollama running the newest open source models on my Nvidia 4070 box, piped to my business machine via the CodeGPT plugin. Nice to have fast, local AI working! qwen2.5 and llama3.2 seem good at writing first draft documentation. Might use to help write the Zerolith docs :O

CodeGPT AI extension in VS code explains and documents Python code in seconds using Llama-3 running locally on your computer (100% free and without internet):

Prompt Injection—A Silent Threat in Developer Extensions Ramiro Molina’s discovery of a prompt injection vulnerability in CodeGPT reveals a critical oversight in how AI extensions are built for developers. The exploit didn’t require elevated privileges or network access—just

太酷了⚡️ 高可用的代码仓库转知识图谱的方案来了🔥 - fork Codex 项目 - 转到 https://t.co/u9YEjaWOsR 创建图谱，等待几秒钟就可以扫码出来 - 打开 VSCode 安装 codegpt 插件 - 在 Copilot 里面将代码库上下文设置好 享受高准确语义级别的代码提问和理解体验💥

Prompt Injection—A Governance Blind Spot The CodeGPT vulnerability disclosed by Ramiro Molina isn’t just a technical flaw—it’s a governance blind spot. No policies were in place to prevent model misuse, and no audit mechanisms existed to detect prompt-based exfiltration.

CodeGPT Prompt Injection—A New Class of Exploit Ramiro Molina’s exploit against CodeGPT demonstrates that prompt injection is more than a novelty—it’s a new class of exploit. By abusing the extension’s prompt handling logic, he was able to extract hardcoded credentials without

CodeGPT now integrates with Meta's Llama 90B, enhancing its code generation capabilities. Explore the latest features! https://t.co/O6YFZV4TE6

Run DeepSeek Models Locally in VS Code 🖥️ Faster, private, and secure AI code assistant with no API keys. Steps needs to do: 1️⃣ Download with Ollama 2️⃣ Install CodeGPT Extension 3️⃣ Go Offline.

I just read Anthropic's guide on building effective AI tools and realized we've been using many of these practices at CodeGPT for months. Here's how we've applied each principle 👇

CodeGPT AI extension in VS Code now has a GitHub AI Agent to help you with all GitHub related topics.

Prompt Injection Exploitation, CodeGPT Ramiro Molina demonstrated how prompt injection can be weaponized against the CodeGPT extension to extract sensitive credentials. This attack vector bypasses conventional security controls by manipulating model behavior through crafted

Say hello to smarter Rust coding with CodeGPT, powered by Gaia 🌱 This tool is built on Gaia’s fine-tuned Rust knowledge model, making Rust code generation and execution seamless and efficient. Integrate CodeGPT into your development setup on the Gaia Network. 🔗 Discover

Vulnerability Disclosure—Prompt Injection in CodeGPT VerSprite researcher Ramiro Molina uncovered a prompt injection vulnerability in the CodeGPT extension that enabled unauthorized access to hardcoded credentials. This was an active exploitation of insecure prompt handling—not a