Skip to main content
← tree
ObjectExpressionskills-ai
const ai & llm engineering: ObjectExpression = {

// LLM integration, RAG, MCP, context engineering, agentic coding, evals, token economics.

Agentic coding tools are systems with mechanics — ReAct loops, context windows, token budgets, subagent spawning — not magic boxes. Understanding those mechanics is what makes the difference between using them and being used by them.

The practical surface: LLM integration and RAG, MCP treated as an architectural primitive rather than a plugin format, prompt and context engineering, and the Anthropic Claude API in production side projects. At work it also means LLM-assisted security triage — running CWE analysis over the output of a scanner built to find chained exploit paths conventional SAST misses, and dispositioning each finding with model-assisted review.

Evals and token economics belong in the same list because neither is optional at scale. A pipeline you can't measure is a pipeline you can't trust, and a pipeline you can't price is one you can't ship.

// 10 keys
LLM integrationRAGMCPClaude APIContext engineeringPrompt engineeringAgentic codingLLM-assisted security triageEvalsToken economics
}
// children (10)
StringLiteralLLM integrationLLM integration · ai & llm engineeringStringLiteralRAGRAG · ai & llm engineeringStringLiteralMCPMCP · ai & llm engineeringStringLiteralClaude APIClaude API · ai & llm engineeringStringLiteralContext engineeringContext engineering · ai & llm engineeringStringLiteralPrompt engineeringPrompt engineering · ai & llm engineeringStringLiteralAgentic codingAgentic coding · ai & llm engineeringStringLiteralLLM-assisted security triageLLM-assisted security triage · ai & llm engineeringStringLiteralEvalsEvals · ai & llm engineeringStringLiteralToken economicsToken economics · ai & llm engineering