LLM Context Engineering Bootcamp

Comprehensive notes and visual resources on LLM context engineering, compiled from multiple sources. The primary resource is the LLM Context Engineering Bootcamp by Vizuara AI Labs (taught by Dr. Sreedath Panat, PhD MIT), supplemented by additional tutorials, papers, and community resources.

What Is Context Engineering?

Context engineering is the discipline of designing, assembling, and managing everything that flows into a large language model's context window to maximize output quality. It goes beyond single-turn prompt engineering to encompass system prompts, memory, retrieval, tool definitions, conversation history, and multi-agent coordination.

Bootcamp Contents

Notes marked with :white_check_mark: are complete. Notes marked with :construction: are in progress.

#	Session	Topics	Status
1	Introduction to LLM Context Engineering	Prompts vs. context engineering, LLM OS analogy (Karpathy), six elements of context, context rot, lost-in-the-middle effect	:white_check_mark: PDF
2	System Prompts & CLAUDE.md	System prompts at the "right altitude", CLAUDE.md / AGENTS.md / skill.md, iterative construction, few-shot example selection	:white_check_mark: PDF
3	RAG from Scratch	WSCI framework, RAG pipeline end-to-end, chunking strategies, hybrid retrieval (dense + BM25 + RRF), cross-encoder reranking	:white_check_mark: PDF
4	Tools, MCP & Agents	Tool schema design, Model Context Protocol (MCP), JIT instructions, ReAct agent loop	:white_check_mark: PDF
5	Context Engineering Failure Modes	Context bloat, context rot, context poisoning, context clash, context distraction, context confusion, context drift, diagnostic framework	:white_check_mark: PDF
6	Multi-Agent Context Management	AGENTS.md, compression, isolation, sub-agent architecture, central brain pattern, fan-out/fan-in, sequential pipelines	:white_check_mark: PDF
7	Memory Management	Episodic / semantic / procedural memory, windowing & summarization strategies, memory decay, cross-session persistence	:construction: In progress

Other contents will be updated in the table as we continue to progress.

Repository Structure

context-engineering-bootcamp/
├── README.md
├── day_01/
│   ├── day_01_notes.typ          # Typst source
│   ├── day_01_notes.pdf          # Compiled PDF
│   ├── brand.typ                 # Shared branding template
│   ├── references.bib            # Bibliography
│   ├── illustrations/            # Slide captures and figures
│   └── mermaid/                  # Mermaid diagram sources + renders
├── day_02/
│   ├── day_02_notes.typ
│   ├── day_02_notes.pdf
│   ├── brand.typ
│   ├── references.bib
│   ├── config-files-comparison.* # Platform config file comparison
│   ├── illustrations/
│   └── mermaid/
├── day_03/
│   ├── day_03_notes.typ
│   ├── day_03_notes.pdf
│   ├── brand.typ
│   ├── references.bib
│   ├── mteb_leaderboard.*        # MTEB embedding model leaderboard data
│   ├── illustrations/
│   ├── mermaid/
│   └── drawings/                 # Hand-drawn diagrams (retrieval, RAG patterns)
├── day_04/
│   ├── day_04_notes.typ
│   ├── day_04_notes.pdf
│   ├── brand.typ
│   ├── references.bib
│   ├── illustrations/
│   ├── mermaid/
│   └── drawings/
├── day_05/
│   ├── day_05_notes.typ
│   ├── day_05_notes.pdf
│   ├── brand.typ
│   ├── references.bib
│   ├── illustrations/
│   └── drawings/
└── day_06/
    ├── day_06_notes.typ
    ├── day_06_notes.pdf
    ├── central_brain_pattern.*
    ├── brand.typ
    ├── references.bib
    ├── illustrations/
    ├── mermaid/
    └── drawings/

Prerequisites

• Basic familiarity with Python
• An API key from Anthropic, OpenAI, or Google Gemini
• Conceptual understanding of LLMs and next-token prediction
• For Day 3: familiarity with vectors, cosine similarity, and a Hugging Face access token

Notes Format

Each day's notes are available in two formats:

Format	Description
.pdf	Compiled document with all illustrations and diagrams embedded
.typ	Typst source — compile with typst compile day_XX_notes.typ

Acknowledgments

The primary source is the LLM Context Engineering Bootcamp by Vizuara AI Labs. Additional content is drawn from various tutorials, research papers, and community resources. These notes are personal study materials created while learning from these sources.