Study series · Vibe Coding & Agentic Engineering

From prompt to production

Five papers, five interactive guides, one goal: master the transition from vibe coding to production-grade agentic engineering. Each guide summarizes a complete paper in interactive form — simulators, quizzes, cheatsheets — in three languages.

📚 5 complete guides🌐 EN · ES · PT-BR🎮 Interactive simulators✅ Quizzes with feedback
Pick a day to start
01
Foundations

The New SDLC with Vibe Coding

The foundation of everything: the spectrum from vibe coding to agentic engineering, the agent loop, context engineering, the factory model, the harness and the token economy.

spectrumagentscontextharnesseconomics
Open guide→
02
Tools & Interoperability

Agent Tools & Interoperability

How agents connect to the world: the N×M problem, MCP and A2A, the orchestrator's role, generative UI and the new payment rails for the agentic economy.

MCPA2Aorchestratorgenerative UIpayments
Open guide→
03
Context & Memory

Context Engineering: Sessions, Memory & Skills

How agents remember and personalize: sessions, events and state; compacting long conversations; the memory pipeline (extraction, consolidation, retrieval) and RAG vs memory.

sessionsmemorycompactionRAGprovenance
Open guide→
04
Security & Evaluation

Vibe Coding Agent Security & Evaluation

How to trust what the agent delivers: quality gates, trust score, adapted DORA metrics, sandboxing, threats like slopsquatting and Denial of Wallet, and production evals.

quality gatestrust scoresandboxingevalsslopsquatting
Open guide→
05
Production

Spec-Driven Production Grade Development

From vibe prototype to production-grade reality: specifications as blueprints, BDD/Gherkin, MCP, agent code review, zero-trust, sandboxing, policy server and team culture.

specsBDDMCPzero-trustculture
Open guide→
📖 About the seriesThese guides summarize and adapt, for educational purposes, the papers of Google's Vibe Coding series. Day 1 lays the foundations; Day 2 covers tools and interoperability; Day 3 deepens context, sessions and memory; Day 4 addresses security and evaluation; Day 5 covers the practices to take agents to production. Each guide is standalone and covers 100% of its paper — but the full journey is better in order.