A practical open-source roadmap to build production-grade AI agents, copilots, RAG systems, MCP tools, and agentic applications.
スター
4
ツール
—
稼働率
—
A premium open-source AI engineering academy for developers who want to build production-grade agents, copilots, RAG systems, MCP integrations, eval pipelines, observability layers, and agentic UI through runnable projects.
Build premium AI engineering skill through runnable systems
Agentic Engineering Playbook is a project-first curriculum for developers, contributors, learners, and recruiters who want evidence of real AI engineering work rather than prompt-only demos.
Learn through six runnable projects
Study production tradeoffs in docs that stay close to implementation
Contribute new chapters, diagrams, examples, and project upgrades
Use the repo as a public proof-of-work system for AI engineering skill
Agentic Engineering Playbook is a practical Docusaurus curriculum focused on production-grade AI application development. It covers the path from direct LLM API usage to RAG, orchestration, MCP, agentic UI, production security, evals, observability, deployment, and commercialization.
an Angular copilot shell for operator-facing AI UX
a safe QA browser agent with local evidence capture
Who should use this
software engineers moving into AI product engineering
full-stack developers building internal copilots
founders validating agentic products
consultants building enterprise AI delivery capability
What makes this different
project-first instead of notes-first
practical enterprise examples
TypeScript-first with Python where useful
security, evals, observability, and deployment included in the main path
designed to create hiring signal, consulting leverage, and product-ready ideas
Why star, watch, or fork
Star the repo if you want a practical open-source AI engineering curriculum to keep improving in public.
Watch the repo if you want updates as new chapters, projects, challenges, and contributor workflows are added.
Fork the repo if you want a personalized learning academy, an internal enablement version, or a portfolio variant tied to your own projects.
Recruiter value
This repository is useful to recruiters and hiring managers because it makes AI engineering skill visible in public:
docs show architecture thinking, production constraints, and system boundaries
projects show runnable implementation skill across backend, UI, evals, and deployment
contribution guides and templates show collaboration readiness
the learning paths make it easy to evaluate focus areas such as Angular AI engineering, agent orchestration, RAG, MCP, and observability
The repo does not claim student counts, adoption, or revenue. Its value is practical proof of work.
90-day roadmap
Days 1-30
finish LLM foundations and provider gateway work
learn prompt contracts, structured outputs, and streaming
publish the first architecture note and demo
Days 31-60
build the enterprise RAG copilot
add retrieval quality thinking, citations, and evals
document tradeoffs, metrics, and failure modes
Days 61-90
build orchestration, MCP, agentic UI, and deployment layers
add approvals, traces, and production hardening
package the strongest projects into portfolio-ready public assets
Repository structure
docs/ Curriculum chapters and roadmap content
projects/ Project briefs and starter implementations
src/ Docusaurus site source
static/ Site assets
templates/ Reusable prompts, checklists, and diagrams
cd projects/p01-ai-provider-gateway && npm run typecheck && npm run build && npm run smoke
cd ../p02-enterprise-rag-copilot && npm run typecheck && npm run build && npm run smoke && npm run eval
cd ../p03-agent-workflow-orchestrator && npm run typecheck && npm run build && npm run smoke && npm run eval
cd ../p04-mcp-enterprise-toolkit && npm run typecheck && npm run build && npm run smoke && npm run eval
cd ../p05-angular-agentic-copilot && npm run build && npm run smoke
cd ../p06-qa-browser-agent && npm run typecheck && npm run build && npm run smoke && npm run eval
CI covers the same safe checks on pull requests and pushes to main. Project 06 deliberately keeps browser-smoke out of CI so the pipeline stays deterministic and does not depend on a browser install or external targets.
Deployment
The repository includes:
CI build validation on push and pull_request
GitHub Pages deployment from main
a Docusaurus config already aligned to AnkitParekh007/Agentic-Engineering-Playbook
If GitHub Pages is not live yet, enable it in repository settings and choose GitHub Actions as the source.
Contributing
Contributions are welcome for curriculum accuracy, code examples, diagrams, Docusaurus improvements, and starter project scaffolding.
最終チェックは 2026-10-28 です。 A practical open-source roadmap to build production-grade AI agents, copilots, RAG systems, MCP tools, and agentic applications.