Comparisons
Honest breakdowns of different approaches to AI context, codebase documentation, and developer productivity.
Cursor Rules vs Intent Nodes
Cursor Rules (Tool-specific context in .cursor/rules/) vs Intent Nodes (Tool-agnostic AGENTS.md context files)
Verdict: Use both. Intent Nodes for portable, hierarchical codebase context. Cursor Rules for tool-specific behaviors and preferences. They complement, not compete.
Documentation vs Codebase Cartography
Traditional Documentation (Human-written docs, wikis, and READMEs) vs Codebase Cartography (Structured context maps for AI agents)
Verdict: Cartography for AI-agent workflows; traditional docs for human onboarding and API reference. Most teams need both, but only cartography compounds.
RAG vs Context Engineering
RAG (Retrieval-Augmented Generation) vs Context Engineering (Structured codebase context)
Verdict: Context Engineering for codebases; RAG for general knowledge retrieval. They can complement each other, but structured context should be the foundation.