835 articles
Script Bash pour Neovim permettant l'intégration d'un assistant IA (Claude) capable de détecter intelligemment le type d'action demandée (remplacement, insertion, amélioration ou affichage) en analysant les instructions de l'utilisateur. Le script lit les fichiers CLAUDE.md globaux et locaux pour appliquer les règles spécifiques du projet lors du traitement des requêtes d'édition de texte.
Claude Code's native projects/ directory provides 30-day transcript retention. Every conversation is automatically saved. This is the raw material.
Different tasks need different specialists. PAI uses a three-tier agent model: TierWhat It IsExampleVoice?Task SubagentsBuilt into Claude CodeEngineer, Architect, Explore, QATesterNoNamed AgentsPersistent identities with backstoriesSerena (Architect), Marcus (Engineer), Rook (Pentester)Yes (ElevenLabs)Custom AgentsDynamic composition from 28 personality traits"Create 5 security researchers"Yes Named agents have their own ElevenLabs voices. When five agents run in parallel, I can hear which one i
The Kai CLI wraps Claude Code with voice notifications. When I run a task and walk away, I hear "Kai starting: Working on research" from across the room. When it finishes, I hear the summary. Every agent has a unique ElevenLabs voice. The Algorithm announces each phase as it executes. This transforms asynchronous work into ambient awareness — I don't have to be staring at a screen to know what's happening. For some people, voice will be the primary interface, not just notifications. They'll talk
He's from the distant future and genderless, but I mostly refer to him as a him. I've named my entire personalized system Kai. Kai is my Digital Assistant that will always be with me, and he is my instance of PAI. Think of it this way: PAI is the architecture—the blueprint for building a Personal AI system. Kai is MY implementation of that architecture, customized with MY knowledge, MY processes, MY domain expertise. He runs on Claude Code today, but the platform doesn't define him—the seven com
Creating a skill is straightforward: Create the directory: ~/.claude/skills/YourSkill/Define triggers in SKILL.md: USE WHEN [keyword patterns]Add workflows: Step-by-step procedures for common operationsBuild CLI tools: TypeScript utilities for deterministic tasks That's it. Claude Code loads all skills into its system prompt at startup. When your request matches a skill's triggers, it routes automatically.
Here's what I find fascinating: PAI, Claude Code, OpenCode, MoltBot—they're all arriving at the same patterns independently. Skills. Hooks. Memory. Agent orchestration. Context priming. This convergence tells us these aren't arbitrary design choices. This is what the architecture of a Personal AI system actually looks like. The seven components aren't my opinion. They're what everyone keeps rediscovering. The question isn't whether these components are right. It's how quickly we can mature each
json{ "mcpServers": { "website-analyzer": { "url": "https://mcp-website-analyzer.YOUR-SUBDOMAIN.workers.dev", "description": "Analyzes websites and extracts key information" } } }12345678
Once connected, you can ask your AI assistant to: "Use httpx to check what technology stack danielmiessler.com uses""Scan these domains for me: example.com, test.com""What security headers does github.com have?" The AI will automatically use the MCP server tools to fetch this information.
🤖 He's from the distant future and genderless, but I mostly refer to him as a him. I've named my entire personalized system Kai. Kai isn't a different AI model or a fork of Claude Code. Kai IS Claude Code—but completely personalized for me. Think of it this way: Claude Code is like macOS or Linux—incredibly powerful out of the box. But just like you customize your operating system with YOUR apps, YOUR shortcuts, YOUR workflows, Kai is Claude Code customized with MY knowledge, MY processes, MY d
🎯 This is THE MOST IMPORTANT part of the entire system. If you take away one thing from this entire post, let it be this: Skills are how you transform Claude Code from a general-purpose assistant into YOUR domain expert. You don't need to fine-tune models. You need to build Skills.
Creating a skill is straightforward: Create the directory structure:~/.claude/Skills/YourSkill/ ├── SKILL.md # Routing and domain knowledge ├── Workflows/ # Step-by-step procedures └── Tools/ # CLI scripts1234Define when to use it in SKILL.md:markdown--- name: YourSkill description: What it does. USE WHEN [trigger phrases] ---1234Create workflows for common operationsBuild CLI tools for deterministic tasksDocument with examples That's it. Now Claude Code knows YOUR
So the umbrella of everything I'm gonna talk about today is what I call a Personal AI Infrastructure (PAI), which is PAI for an acronym. Everyone likes pie. It's also one syllable, which I think is an advantage. But here's what makes PAI v2 different from what came before: This isn't just a collection of prompts and tools anymore. PAI v2 is about taking Claude Code—which is already incredible—and personalizing it into YOUR cognitive operating system. Think of Claude Code as the foundation, like
The best prompt engineering is building a system that doesn't need perfect prompts. Everyone's obsessed with prompt engineering. "How do I write the perfect prompt?" "What's the magic phrase that makes GPT-4 work better?" The real power is in building a system that gives YOUR AI the context, tools, and structure to understand what you actually want, even when your prompts are messy. Instead of spending 20 minutes crafting the perfect research prompt every time, you build a Research skill once th
📁 Context lives in Skills, not a separate directory. Context isn't stored in a separate "context" directory anymore. Context IS the Skills. Each Skill contains its own knowledge files: ~/.claude/Skills/Blogging/ → Blog writing standards, style guides~/.claude/Skills/Research/ → Research methodology files~/.claude/Skills/Art/ → Visual aesthetic guidelines All Skills are pre-loaded into Claude Code's system prompt at startup. When you make a request, the routing system matches it to the appropria
🪝 Hooks are the nervous system—they connect everything automatically. We've covered Skills (WHAT to do), Context (WHAT to know), and History (WHAT to remember). Now let's talk about WHEN things happen automatically. The Hook System makes your personalized Claude Code reactive.
SessionStart Hook (~/.claude/hooks/session-start/): typescript// When session starts: // 1. Load CORE context (identity, principles, contacts) // 2. Check for active tasks from previous sessions // 3. Initialize observability tracking // 4. Set up voice server connection12345 Every time you start Claude Code, this runs automatically. You don't ask for it. It just happens. Stop Hook (~/.claude/hooks/stop/): typescript// When session ends: // 1. Extract 🎯 COMPLETED message from final response //
At the center of my workflow is a custom kai command that wraps Claude Code with voice notifications and context management. bash# Interactive mode (default PAI directory) kai # Single-shot query with voice kai "what's my schedule today?" # Run with specific context directory kai --context ~/Projects/Website "deploy the latest changes" # Silent mode (no voice) kai --no-voice "analyze this code for security issues" # Wallpaper management (Kitty terminal integration) kai wallpaper circuit-boar
Since ChatGPT arrived in late 2022, we have been bullish on what large language models could do for reverse engineering and malware analysis. The early models were useful for teaching but too rudimentary for production work; that changed with the advent of reasoning models. OpenAI’s o1-preview, in September 2024, was the first to show the kind of sustained problem-solving the work demands, and within months Sean Heelan had used o3 to find a net-new vulnerability in the Linux kernel. In cybersecu
A developer with a good AI assistant can produce more and produce faster, but ten developers all doing that, each in their own way, with their own tools and their own context, don’t add up to a team that is ten times better. More often they add up to a team moving faster in ten different directions. The speed stays with the person who created it. The reasoning, context, and decisions that the rest of the team would need to build on that speed gets lost. These three structural problems explain wh
We are delighted to announce the introduction of a free, open-source MCP Server and Agent Skill for Bulk Export. Bulk export is a highly efficient way to access all your Rapid7 vulnerability and exposure data to AI assistants and custom AI workflows. Built as an open-source bridge, it helps customers bring their Rapid7 data into the tools and experiences that work best for their teams. Check out our blog for more detail. Figure 3: Agent Skill for Bulk Export.
Traditional web applications have defined operations: check a bank balance, make a transfer. You can write deterministic rules to secure those interactions. AI-powered applications and agents are different. They accept natural language and generate unpredictable responses. There's no fixed set of operations to allow or deny, because the inputs and outputs are probabilistic. Attackers can manipulate large language models to take unauthorized actions or leak sensitive data. Prompt injection, sensi
Overall, there is too much data in engagement logs for anyone to manually review, causing a bottleneck for analysis. To combat this issue, Janus produces normalized JSON for every task and result, making operational data usable by LLMs. Users should ensure any data shared with LLM services aligns with their organization’s data handling policies. For sensitive environments, this can be done using self-hosted models or hosted options such as Amazon Bedrock. Instead of digging through logs, teams c
Before rappelling into the depths of application code, it is beneficial to provide Claude Code with contextual information for the target application and guidance on what we want to see in a response. This will result in fewer false positives and more useful responses compared to just telling Claude to “Find vulns in this app plz”. A system prompt is the perfect place to supply this. When working with an LLM, the client application maintains a rolling, limited “memory” of your conversation, call