All Vaults
CE

Context Engineering

Prompts · MCP · Agent Context

The emerging discipline of designing what an LLM sees — prompt, tools, MCP, memory, retrieved data — so agents can complete long-horizon tasks reliably.

20 articles·0 phases·Updated 4/21/2026·
Curated byburn451
Get Burn 451

The emerging discipline of "context engineering" — designing what an LLM sees (prompt + tools + MCP + memory + data) so it can complete long-horizon tasks reliably. Term popularized by Harrison Chase (LangChain) in June 2025, embraced by Anthropic, Simon Willison, and the broader agents community. Spans from MCP specification and tool schemas to long-context strategy and agentic harness architecture.

About this vault

Context engineering is how AI engineers build agents that actually work. This vault curates 20 canonical pieces spanning three axes: (1) the Model Context Protocol — specification, Anthropic's announcement, Thoughtworks' impact analysis; (2) the discipline itself — Harrison Chase's 'rise of context engineering', LangChain's scratchpad/memory patterns, Philipp Schmid's breakdown; (3) practitioner voices — Simon Willison, Andrej Karpathy, and the Sequoia podcast with Harrison Chase on long-horizon agents. The organizing thesis: prompt engineering gets you a clever reply; context engineering gets you a reliable agent. Curated 2026-04 as the term crystallized from Twitter meme into engineering practice.

20 articles

MCP: The Protocol Layer

Model Context Protocol turned 'tool use' from ad-hoc integrations into a real standard. Anthropic's announcement, the spec, Thoughtworks on MCP's 2025 impact — the plumbing that made context engineering portable.

Agentic Harness & Long-Horizon Architecture

Context engineering at the system level — agentic harnesses, agent engineering, scratchpads, memory tiers, and the Sequoia podcast with Harrison Chase on what it takes to keep agents coherent across long tasks.

Practitioner Voices

Simon Willison, Andrej Karpathy, Philipp Schmid, and others on what actually matters in context design. Sharper, less framework-y takes from people shipping with these models daily.

Start reading, not hoarding.

Import this vault to Burn 451 and actually read what matters.

Frequently asked questions

Who is Context Engineering?

Context Engineering is covered in this Burn 451 vault with a focus on prompts · mcp · agent context. The emerging discipline of designing what an LLM sees — prompt, tools, MCP, memory, retrieved data — so agents can complete long-horizon tasks reliably.

How was the Context Engineering vault curated?

The Context Engineering vault was hand-curated by the Burn 451 editorial team from publicly available essays, blog posts, podcast transcripts, and social threads. Each piece includes an AI-generated summary so readers can triage in seconds. The vault auto-syncs as new content from Context Engineering is published.

How many articles are in the Context Engineering vault?

The Context Engineering vault currently contains 20 curated pieces organized by topic, not chronology. Each article has an AI summary and a direct link to the original source. Items are refreshed hourly through Burn 451's ISR pipeline, so new publications appear within a day.

How do I use this vault with Claude or Cursor?

Install the burn-mcp-server package from npm and connect it to Claude, Cursor, or any MCP-compatible AI tool. The vault becomes queryable as live context — your AI can search, summarize, and cite articles from Context Engineering directly in conversation without manual copy-paste or re-uploading files.

What is Burn 451?

Burn 451 is a read-later app built around a 24-hour burn timer that forces daily triage. Articles you save must be read, vaulted, or released within 24 hours. The Vault layer — including this Context Engineering collection — holds permanent curated reading lists for AI thought leaders, founders, and researchers.

Content attributed to original authors. Burn 451 curates publicly available writing as a reading index. For removal requests, contact @hawking520.