Reading List
The framework is a synthesis; these are the primary sources. Start with the practitioners, then the official engineering writing, then the maturity-model literature.
Practitioners
Section titled “Practitioners”- Boris Cherny — Best practices for Claude Code · How Boris Uses Claude Code
- Andrej Karpathy — “Software Is Changing (Again)” / Software 3.0 · the original “vibe coding” framing and its 2026 “agentic engineering” retrospective
- Thariq Shihipar — Claude Code is All You Need · “HTML is the new markdown”
- Simon Willison — Agentic Coding
- Armin Ronacher — Agentic Coding Recommendations
- Geoffrey Huntley — Ralph Wiggum as a software engineer
- Thorsten Ball — How to Build an Agent
- Hamel Husain — Your AI Product Needs Evals
- Steve Yegge — Revenge of the Junior Developer
Official engineering writing
Section titled “Official engineering writing”- Building effective agents — the workflow-vs-agent taxonomy and the five workflow patterns
- Effective context engineering for AI agents
- Writing effective tools for AI agents
- How we built our multi-agent research system
Maturity-model & metrics frameworks
Section titled “Maturity-model & metrics frameworks”- Capability Maturity Model — the 5-level shape
- DeepMind: Levels of AGI — performance × generality, with autonomy as a separate axis
- SAE J3016 levels of driving automation — division of responsibility as the basis for levels
- Factory.ai: Agent Readiness — the binary, file-existence codebase rubric
- DORA 2025 AI Capabilities Model — the team/org layer
This list grows as new high-signal work is published and folded into the framework.