AI & LLMs

The fastest-moving domain in the brain. Curated concepts, patterns, and tools at the intersection of artificial intelligence and software engineering.

Claude Code

PageDescription
OverviewAnthropic’s agentic CLI tool for repository-level engineering
Plan ModeStructured planning before execution — reduces errors on complex refactors
Custom Slash CommandsExtending Claude Code with domain-specific workflows
SkillsReusable prompt templates and skill definitions
SubAgentsIsolated child agents for bounded task execution
Custom Subagent PatternsFile format, frontmatter reference, and canonical pattern library

Workflows & Patterns

PageDescription
Agentic WorkflowsIterative, tool-augmented reasoning and multi-agent coordination
AI-Assisted WorkflowsAutonomous AI assistant, Context7, and Tavily knowledge stack
Spec-Driven DevelopmentUsing specifications to drive agentic implementation
Vibe CodingRapid, intuitive AI-assisted development philosophy

SDLC

PageDescription
Agent-Native SDLCThe “Software Factory” pattern and upstream context engineering
AI-Augmented SDLCEvolution of DevOps where agents participate in the full lifecycle
Agentic DevOpsAgents in CI/CD, incident response, and infrastructure management

Infrastructure

PageDescription
MCPModel Context Protocol — standards for AI-to-tool communication
Top 10 MCP ServersCurated production-ready MCP server integrations
RAGRetrieval-Augmented Generation architectures
Agent Memory SystemsShort-term, long-term, and episodic memory architectures
Context EngineeringVersioning and testing of agentic prompts and data
Fabric FrameworkModular human-augmentation framework using Patterns and Stitches
Minimalist Agent DesignThe “pi” philosophy of transparency and tool minimalism

Observability & Research

PageDescription
LLM ObservabilityMonitoring token usage, latency, and context integrity
GPT-5.5 vs Claude 4.7Technical benchmark on Cloudflare Workers optimization
DPO InsightsDirect Preference Optimization research findings

Career

PageDescription
AI Career TransitionCore differences from traditional programming and auxiliary paths

Tags: ai llm claude agentic devops workflows infrastructure observability