https://lucaslinkowski.com/blog.html Technical writing on malware analysis, reverse engineering, detection engineering, and GenAI in cybersecurity. en-gb https://lucaslinkowski.com/blog/agent-harnesses-skills-and-frontier-model-architecture.html A long form guide to agent harnesses, skills, MCP servers, tool runtimes, and the system design choices that turn frontier models into real task systems. Tue, 28 Apr 2026 09:00:00 GMT https://lucaslinkowski.com/blog/agent-harnesses-skills-and-frontier-model-architecture.html https://lucaslinkowski.com/blog/weaponizing-web-lures-against-agentic-ai.html How cybercriminals are adapting SocGholish and ClickFix campaigns to target autonomous AI agents through Indirect Prompt Injection and the Parallel-Poisoned Web. Sat, 28 Mar 2026 09:00:00 GMT https://lucaslinkowski.com/blog/weaponizing-web-lures-against-agentic-ai.html https://lucaslinkowski.com/blog/mcp-server-malware-traffic-analysis.html How I built a natural language interface connecting Fiddler traffic capture to Gemini LLM using the Model Context Protocol. Fri, 21 Mar 2026 09:00:00 GMT https://lucaslinkowski.com/blog/mcp-server-malware-traffic-analysis.html https://lucaslinkowski.com/blog/detecting-ai-prompt-injection-in-the-wild.html Detection patterns for indirect prompt injection, AI ClickFix for computer-use agents, CSS hidden text traps, HashJack fragments, and credential exfiltration targeting autonomous AI agents. Sat, 15 Mar 2026 09:00:00 GMT https://lucaslinkowski.com/blog/detecting-ai-prompt-injection-in-the-wild.html https://lucaslinkowski.com/blog/cognitive-induction-prompting.html A deep dive into the 5-step LLM generation pipeline and how understanding tokenisation, embeddings, attention, prediction, and sampling transforms prompting from guesswork into deliberate control. Mon, 10 Mar 2026 09:00:00 GMT https://lucaslinkowski.com/blog/cognitive-induction-prompting.html