Hi, welcome to my garage 👋

Agentic AI & Security Engineer · Paris / Learn, Build, Share and Connect

Website · Email · GitHub · LinkedIn · X

👨‍💻 About Me

Software engineer in Paris with 15+ years of experience, 10+ of them in network security at Montimage. At work I lead development of MMT-DPI, a deep packet inspection engine written in C, and build LLM pipelines for security: fine-tuned small models, LLM-as-judge evaluation and local deployment with Ollama. In my own time I build tools for AI agents: guides, skills, MCP servers and usage analytics, mostly around Claude Code. I build them with Intention-Driven Development (IDD), my workflow for turning GitHub issues into work orders an agent can ship.

I value clarity, reliability, and measurable impact over complexity.

🛠️ What I'm Building

Project	What it does	Stars
claude-howto	Visual, example-driven guide to Claude Code, from first prompt to agents, hooks, skills and MCP
IDD Stack	Intention-Driven Development: eight skills that plan, triage, analyze and resolve GitHub issues, turning each into a tested PR (website)
asm	CLI to install, audit and scan agent skills across 21 coding agents
skills	Reusable skills for AI coding agents
context-stats	Claude Code usage analytics: context window, cache efficiency and cost
mitre-mcp	MCP server giving AI assistants MITRE ATT&CK knowledge (official MCP Registry)
sec-mcp	Threat-intelligence checks for domains, URLs and IPs, as a CLI or MCP server

👀 Profile Views (Self-Hosted - sometime it has broken because number of visit, i am on free plan only)