Containerlab sFlow-RT Development Environment

Sometimes the best tools are the ones you build yourself. 🛠 . I built this customized Windows Manager from scratch vibe codedly. . No installs. No bloat. Pure Python. . Here's what it does: 📌 Pin any window to always stay on top 👁 Control opacity per window — slide any app from solid to ghost ✕ Close windows directly from the list 🎨 8 built-in themes (Dark, Nord, Amber, Rose Pink + more) 🔄 Live refresh — see all your open windows in real time . More updates to follow.

My agents kept timing out when calling Linear. I realized MCP was loading every tool schema with each message, which meant 25K to over 60K+ tokens per round trip before anything useful happened. I switched to Python scripts that called the API directly, which reduced token usage to 1K-2K per call. The scripts were messy, though. So I rewrote everything in Rust. That’s what Lin is now. https://lnkd.in/g9X5hDHf                                           $ brew install aaronkwhite/tap/lin $ cargo install lincli Using Linear? Try it out. Still tweaking things so feedback's genuinely useful. #DeveloperTools #AIAgents #OpenSource #Linear #MCP #API #RUST

I built a VS Code extension to fix a real problem I kept hitting in Claude Code. When working with AI, documentation is everything… but Claude Code can’t reliably access full docs on its own. So I built DocCrawler MCP Server VS Code Extension: https://lnkd.in/gFUurv8z What it does: Crawl entire documentation sites from a single URL Clean and structure the content (no navbars, sidebars, junk) Feed it directly into Claude Code via MCP Works even with JavaScript-heavy docs (SPAs) Why I built this: I was constantly copying docs manually into Claude Code… which is slow and breaks the whole “AI workflow” idea. Now I just run: fetch_docs(url="docs site") …and Claude Code instantly has the full context. Or you can just give a prompt and tell claude code to use mcp server fetch_doc for website url. Key Features: Seamless MCP integration with Claude Code Auto setup (Python + Playwright installs automatically) Simple .mcp.json config Cross-platform (macOS, Windows, Linux) Impact: This makes Claude Code actually useful for real-world dev work - no more partial context, outdated answers, or manual copy-paste. If you’re using Claude Code, this will save you a ton of time. Would love feedback or ideas

10 Examples, Pipe, Linux ls | sort (files icon) ls | wc -l (counter icon) cat file.txt | grep hello (search icon) head file.txt (top lines icon) tail file.txt (bottom lines icon) ps aux | grep python (process icon) sort file.txt | uniq (filter icon) cat file.txt | wc (document icon) du -h | sort -h (storage icon) grep error file.txt | wc -l (error icon)

🧠 Schedule Your Python Scripts to Run Automatically — No IT Team Needed You’ve written the script. Now what — do you run it manually every Monday? No. You schedule it. 🔹 On Windows (Task Scheduler) Create a simple .bat file: python C:/scripts/weekly_report.py Schedule it once → it runs automatically every week. 🔹 On Mac/Linux (cron job) 0 8 * * 1 /usr/bin/python3 /scripts/weekly_report.py This runs every Monday at 8 AM. 🔹 On Cloud Use tools like: 👉 GitHub Actions 👉 AWS Lambda Even the free tier is often enough for basic automation. 💡 The real shift Automation isn’t just writing the script. It’s making the script run itself. That’s where: 👉 time savings compound 👉 manual work disappears 👉 processes become reliable The best scripts are the ones you never have to run manually again. 👉 What’s one script you could schedule today? #ActuaryWhoCodes #PythonForActuaries #Automation #Productivity #DataAnalytics #Analytics

Writing the script is step one — scheduling it is where the real value comes in. #PythonForActuaries #Automation #Productivity #DataAnalytics #Analytics

How exactly do you install Convai Agent? We absolutely hate tools that take 40 complicated steps to install. You clone a repository, install endless python dependencies, manually fix path variables, and after three hours, it still breaks on Windows. We wanted this to be radically different. We built a native Electron application specifically optimized for Windows. You just download the setup file, run the installer, and you are completely done. The native C++ daemon, the secure local folder bridge, and the beautiful Terminal UI are all packaged together. On the very first launch, it automatically downloads the necessary GGUF models for local execution. And because we utilize a Hybrid API approach, you can instantly connect to our cloud endpoints for heavier tasks without configuring any complex server infrastructure. Built for clinical scientists who need to focus entirely on their research, not on debugging Python environments. Experience the seamless setup with Convai Innovations. https://lnkd.in/gH-9gBfT

Running a local LLM usually means a Python environment, CUDA drivers, and at least one Stack Overflow tab open before you’ve even started. llamafile skips all of that. Mozilla.ai packaged the whole runtime like model weights and everything into a single executable. On Windows you rename it to .exe. On Mac or Linux you chmod +x it. That’s the setup. https://lnkd.in/gn6mAyz3 #opensource #llama #ai #llamacpp #artificialintelligence #gguf #huggingface

Containers can feel very reliable… until they're not. One thing I have seen more times than I can count: An application works perfectly on a developer's laptop, but once it's inside a container, something breaks. Most times, it is not Docker itself that is the problem. Here is what it actually ends up being: 1. Missing Dependencies Your local machine has Node, Python, or a system library installed globally. The container does not. The app runs locally but fails in the container because that dependency was never declared. 2. Environment Variables Your .env file works on your machine, but you forgot to pass it to the container. Suddenly the app cannot find the database connection string or API key. 3. File Paths Windows uses backslashes. Linux uses forward slashes. Your container runs Linux. That hardcoded path C:\projects\data will not work. 4. Assumptions About the Runtime Environment You assumed Python 3.10 is installed. The base image uses 3.8. You assumed /tmp is writable. Maybe it is mounted read-only. Containers force you to be explicit about everything. And that is a good thing. It exposes hidden assumptions and makes your application more portable and reproducible. But only if you pay attention to the details. Here is what I do now: · Always build from a clean base image locally before pushing · Explicitly list every dependency in the Dockerfile · Pass environment variables intentionally, never by accident · Use relative paths or environment-specific path variables · Test the exact same image in staging before production The more predictable your container is, the more reliable your system becomes. #Docker #Containers #DevOps #CloudComputing #AWS #ECS #TheEmpatheticEngineer

Hyperparameter Optimization Machine Learning using dragonfly #machinelearning #datascience #hyperparameteroptimization #dragonfly Dragonfly is an open source python library for scalable Bayesian optimisation. Bayesian optimisation is used for optimising black-box functions whose evaluations are usually expensive. Beyond vanilla optimisation techniques, Dragonfly provides an array of tools to scale up Bayesian optimisation to expensive large scale problems. These include features/functionality that are especially suited for high dimensional optimisation (optimising for a large number of variables), parallel evaluations in synchronous or asynchronous settings (conducting multiple evaluations in parallel), multi-fidelity optimisation (using cheap approximations to speed up the optimisation process), and multi-objective optimisation (optimising multiple functions simultaneously). Dragonfly is compatible with Python2 (>= 2.7) and Python3 (>= 3.5) and has been tested on Linux, macOS, and Windows platforms. https://lnkd.in/gPQUfsR9