Derniers Articles

The memory system behind an Agentic OS
Thursday, May 7, 2026

Not one AI memory — three problems: Storage, Retrieval, Consolidation. 6 levels broken down for Claude Code, from CLAUDE.md to sleep-time compute.

AI development in companies: what my interview with Alex Soyer reveals
Tuesday, May 5, 2026

On parle beaucoup des devs qui automatisent tout avec Claude Code. Mais en entreprise, le sujet est plus complexe : sécurité, conformité, SOC2, culture interne, responsabilité et d

cmux: the terminal built for agentic development
Tuesday, May 5, 2026

Quand tu lances un seul agent Claude Code, ton terminal classique suffit. Mais dès que tu lances plusieurs agents en parallèle, avec plusieurs worktrees et plusieurs tâches longues

Claude Design tested on a real SaaS onboarding
Tuesday, May 5, 2026

Claude a longtemps été faible sur le design produit. Avec Claude Design, Anthropic essaie de combler ce trou : passer d’un assistant qui raisonne sur du code à un assistant capable

Claude Mythos and zero-days: why developers need to wake up
Tuesday, May 5, 2026

Anthropic a montré que Claude Mythos pouvait trouver de vraies failles dans du code critique. Le sujet n’est pas juste la performance du modèle, mais le changement d’époque pour le

The memory system behind an Agentic OS
Tuesday, May 5, 2026

Claude Code a maintenant une mémoire utile pour un projet. Mais construire un Agentic OS comme OpenClaw demande une vision plus large : plusieurs projets, plusieurs canaux, plusieu

GStack: the Y Combinator pack for Claude Code
Sunday, May 3, 2026

GStack turns Claude Code into a team-like agent system with roles, checklists and workflows inspired by Y Combinator. Here is my test on a production SaaS.

Token Anxiety, AI Burnout and Disconnecting: What One Week Off Taught Me During the Opus 4.7 Launch
Monday, April 20, 2026

Why one week off in Thailand forced me to rethink token anxiety, AI burnout and the pace created by launches like Opus 4.7.

Alex Soyer Interview - The Real State of AI Dev in 2026
Thursday, April 16, 2026

58 minutes of honest discussion about OpenClaw, Claude Code, SOC2, SaaS, and the growing gap between solo devs and enterprise teams.

Anthropic Mythos and the End of Security Complacency
Friday, April 10, 2026

Why Claude Mythos changes the cybersecurity conversation for developers and engineering teams.

Don't buy a Mac Mini for OpenClaw before seeing these benchmarks
Thursday, April 9, 2026

I compared a MacBook Pro M4 Max, an RTX 5080, and a VPS to run OpenClaw in real conditions. The results are far more nuanced than what most local AI demos suggest.

Anthropic just killed OpenClaw (but there is still a solution)
Monday, April 6, 2026

Anthropic blocked OAuth access for third-party tools. Here are the 3 ways to keep using OpenClaw with Claude, and the deeper lesson about resilient AI workflows.