Rudolf Groetz’s Post

Agent performance isn’t just ‘more tools’. Manus shows why context engineering wins: VM sandbox execution, KV-cache discipline, masking actions, and file-based memory.

Step 1: Create a problem by inventing a language that nobody needed. Step 2 : Create another problem by creating a tool that nobody needed. Step 3 : Solve all those problems by adding just another tool. 🤷🏻♂️🤷🏻♂️🤷🏻♂️

It’s time for Part 4 of ctrl+shift+speed! 🚀 In this series we share helpful ISG-virtuos shortcuts that cut down clicks and speed up simulation workflows. Last time we revealed how to position elements automatically. Still remember how? Here’s a quick refresher: press + hold “alt” while the object is selected to activate temporary geometry snapping. Got it? Perfect, then you’re ready for today’s tip. 💡 In this part we’ll show you a faster way to link ports or blocks in block diagram mode. Select the port or block to start from. Press and hold “Shift,” and click the port or block you want it to connect to or draw a selection frame to connect multiple ports. With just one click, the first free matching inputs and outputs are connected automatically. Which shortcut saves you the most time? Let us know in the comments 👇

Vulnerability scanning hasn't kept up. Most tools still grade you on "is it the latest version?”, which is how you end up with PDFs full of "critical" maybes and a frustrated engineering team. We take a different path: validate exploitability at runtime and return evidence for every finding, to get to yes/no instead of maybe. We break it down with a CrushFTP example and show how Nuclei's template/flow engine turns noise into action. Read the blog → https://lnkd.in/gJbBnKqW #vulnerabilityscanning #vulnerabilitymanagement

Just published my first blog at ProjectDiscovery! I dove into why vulnerability workflows keep breaking between detection and validation and how we can finally fix them for good. If you’ve ever stared at a “critical” finding that turned out to be a “maybe,” this one’s for you.

Source: https://lnkd.in/dXMpehzB 🚀 Immutable Design Meets Performance💡 Ever wondered how immutable systems can stay fast? This deep dive shows that memoization + HashSet can turn complex task scheduling into a breeze! 🚀 Cache results to avoid redundant work in Expand. Use HashSet for lightning-fast prerequisite checks. Balance immutability with internal mutability for speed. 🔧 Engineers: Re-evaluate your assumptions about immutability—performance isn’t mutually exclusive! 💡

In this quick demo, Luca Dennig walks through how to use a trigger event to control the rotation of a door on a box being filled with water. As the water level rises past a defined threshold, the door automatically opens to prevent overflow. The setup includes locking the door’s rotation at the start of the simulation, tracking the fluid level inside the box, and defining when and how the door should respond. It’s a clear example of how shonDy supports event-driven logic - letting engineers simulate realistic responses to physical thresholds, without scripting workarounds or manual interventions. Watch the video to see how it works in action! Explore shonDy: https://lnkd.in/egM2J4KX #SimulationTools #TriggerEvents #CFD #DigitalEngineering #FluidSimulation

#hardhatINSIGHT: Success in the Built Environment is now defined by quantifiable outcomes like Whole Life Carbon (WLC) and physical resilience. But how do we achieve the Fused Design methodology required to meet these rigorous standards? READ HERE: https://lnkd.in/d34fKNdQ

How long should validation take? NoIdea ideas include realistic timelines. Example progression: → Week 1: Build promise extraction (test on 50 samples) → Week 2: Deploy survey mechanism (60-second completion) → Week 3: Score responses (80% reliability threshold) → Week 4: Test with pilot cohort (20 teams) Not "eventually." Specific weeks with specific goals. That's lean validation. Join the waitlist 👉 https://noidea.one

FlowImprove subroutine is designed to enhance an existing graph partition by minimizing a quotient cut score. It achieves this by iteratively refining the partition using a series of maximum flow computations. The algorithm operates on a weighted graph and adjusts the partition to achieve a better balance between the cut size and the partition sizes. 💡 In short: FlowImprove’s key advantage is that it takes an existing partition and makes it significantly better using precise, max-flow-based refinements, all while keeping the partition balanced, something most heuristic methods cannot guarantee.