Vibecoding EN Subscribe

Vibecoding.cz — English — Claude Code and AI coding articles

Patrick Zandl on Claude Code, coding agents and the AI developer-tools market.

Latest · Tutorial

How Claude Code Handles Memory About Your Project

You correct Claude on the same thing for the second time and wonder whether it will remember it this time. Claude Code now has two layers of persistent memory: rules that you write, and notes that it keeps on its own. Both are plain Markdown files on your disk. That is the biggest advantage of the whole approach and also its weakness, because a note in a file only helps if Claude remembers to open it. How do you work with this so that you end up satisfied?

Nov 2, 2026 · 8 min read

This week in Antigravity CLI Paste or drag images straight into the prompt, plus hooks that can stop a turn All weekly updates →
This week in Claude Code ultrathink now runs at xhigh effort, and subagent effort has its own setting All weekly updates →
This week in Codex CLI Google Cloud Vertex AI arrives as a built-in provider All weekly updates →
This week in Grok Build Experimental vim editing in the prompt, plus fail-closed hooks All weekly updates →
This week in Muse Code Built-in skills for 3D pages and browser games that test their own output All weekly updates →

Recent

Article Gemini 4 Carbon has caught up with Opus, but you can't use it yet After a year, Google is back at the top of the independent Artificial Analysis leaderboard, and on the Vals Index it matches Claude Opus 5.5 at a third of the price. If you want to try it right now, you're out of luck: for the time being, Carbon goes only to vetted defenders in cybersecurity. Where does it actually fit in Google's lineup, and what should you do until it becomes available? Oct 29 · 6 min read Article GPT-6.1 Terra only pays off at half throttle Ten times pricier per token than Claude Haiku 5.5 and four times pricier than DeepSeek V4.1 Flash. Still, at medium effort Terra delivers more than both rivals at maximum, for less than half the cost per task. The catch is how you run it. From xhigh upward, the same company's GPT-6.1 Sol is both better and cheaper. Oct 28 · 9 min read Tutorial Fewer Subagents, or When Too Many Agents Hurt in Claude... It's tempting. You split an agent into specialists, each team takes its own piece, and an orchestrator ties it all together at the top. Finally, dozens of agents are working for you! It looks clean, it grows along with the organization, and everyone knows what they're responsible for. Based on its experience with enterprise deployments, Anthropic claims that such a setup often performs worse, runs slower, and costs more than a single agent with Skills. So when should you merge subagents, when should you keep them, and how do you tell that merging didn't break something? Oct 26 · 10 min read Article Kimi K3.1 Pro: the best open model is slow and not as cheap as it looks Three points behind Claude Opus 5.5, downloadable weights, and half the price per token. On paper, it's the best open model ever released. In practice, though, it responds slowly and thinks a lot. Compared to Opus, depending on the task and settings, you save two thirds or nothing at all, and compared to GPT-6.1 Sol or DeepSeek it's expensive. So who is it actually worth it for? Oct 24 · 7 min read Claude Code Claude Duet: The Assistant That Reaches Out on Its Own. Once Every Few Hours It's tempting, I'll admit. In the morning your assistant tells you what's on fire, during the day it watches your email, your calendar and your unpaid invoices, and it only pings you when it needs a decision from you. Anthropic has now released exactly that kind of agent (via a waitlist) into a field of several souped-up competitors. What does it promise? And what does it actually do? Oct 23 · 10 min read

One email when something worth reading ships.

Every Tuesday: what changed in Claude Code, Codex and other coding agents, plus new articles. No spam.

From the archive

  1. Oct 22 The First Exploit of the AI Agent Era Agents & security
  2. Oct 20 Add-ons that promise to make Claude Code cheaper: I measured them on 84 runs Claude Code
  3. Oct 13 When Harari Talks About AI After a Booed Macinka Harari
  4. Oct 5 Decision Models: Cheap AI That Picks the Right Answer Tutorial
  5. Oct 2 Mods for Claude Code: Anthropic Lets You Under the Hood Claude Code
  6. Sep 30 OpenAI DevDay 2026: Dots Personal Agents and Other News OpenAI & Codex
  7. Sep 29 The $200 Codex Pro Plan Is Back. With Half the Volume OpenAI & Codex
  8. Sep 21 Who profits when you run an LLM at home Local AI
  9. Sep 19 Jev: the model that doesn't write, it just decides. How to use it and when it pays off Tutorial
  10. Sep 18 Claude Code Projects Now Split Up the Work on Their Own Claude Code
  11. Sep 16 Černá AI Festival Lit Up Ostrava events
  12. Sep 14 How to Tell a Good AI Workshop from a Bad One Workshops
  13. Sep 13 Cursor Projects Shows a Great New Approach to Managing Agents Cursor
  14. Sep 9 Anthropic's head of AI safety testing: We don't have a plan yet for aligning superintelligence Claude Code
  15. Aug 31 OpenClaw 2.0 Does Multiplayer. Watch Out for It! Agents & security
  16. Aug 28 The Case of OpenAI's Rogue AI Agents Attacking Hugging Face OpenAI & Codex
  17. Aug 26 Hardware for Local AI: What to Buy in August 2026, and Whether It Pays Off Local AI
  18. Aug 12 An AI watermark proves touch, not authorship. What should we do about labeling AI content? Claude Code
  19. Aug 11 The Agent Deleted Someone Else's Booking: What AI Agents Do to a Sloppy SaaS Agents & security
  20. Aug 10 Jeff Dean Is Leaving Google. He Has Already Become a Legend. Google
  21. Aug 7 Grok Build Reaches Version 1.0 Grok
  22. Aug 6 Meta Muse Code: When Zuckerberg Teaches AI to Code Meta
  23. Jun 3 Claude Code rewrote /fork. For the second time this month. And this time, it was useful Tutorial
  24. May 31 Dynamic Workflows in Claude Code: An Announced Experiment Now Set to Be Put to the Test Tutorial
  25. May 16 Grok Build CLI – xAI Enters the Terminal Coding Agent Market Grok
  26. May 14 Claude Code Agent view—it’s a whole new way of working Claude Code
  27. May 12 Claude Code also offered /goal for automatically reaching a destination Claude Code
  28. Jun 2, 2025 🐑 Sheep Marketplace — how Tomas Jirikovsky and the Czech Minister of Justice came to Bitcoin Bitcoin
  29. May 14, 2025 Claude Code Hooks — The Complete Guide Tutorial
  30. Mar 7, 2024 When AI comes for those with political power… AI
  31. Oct 13, 2023 The hope of the year: the RISC-V processors RISC-V
  32. Jul 10, 2017 First Czechoslovak radiotelephone networks: AMR Telecom History
  33. May 11, 2016 Strategy of motivation to change the behavior in the wearable and Internet of Things world —… Wearables
  34. Feb 10, 2016 The fairy tale about a single standard for the Internet of Things Internet of Things
  35. Jan 21, 2016 Edge Computing and Fog Computing; or, on the Benefits of Fogging Edge Computing
  36. Jan 4, 2016 WTF of Energy Efficiency: A Quest for the Right Metrics Internet of Things
Patrick Zandl

Patrick Zandl runs Claude Code workshops in Prague and writes about what actually works in AI-assisted software projects, and what doesn’t. This is the English edition of Vibecoding.cz, which covers a broader range of AI tools in Czech.