Concept detail
Context engineering
12 tagged videos · 16 connection videos · degree 0
Publication timeline
- 2026-021 video
- 2026-031 video
- 2026-051 video
- 2026-083 videos
- 2026-0912 videos
- 2026-103 videos
Source videos
Observed neighbors
12-Factor Agents
context window ownership and structured outputs are concrete instantiations of the context-engineering discipline
1 source videos
Ex-NASA dev reveals his Agentic Engineering Workflowartifact chain (intent.md/spec.md/plan.md)
Externalizes and versions the context supplied to agents, replacing chat history with reviewed repository artifacts.
1 source videos
Claude Code: The Complete AI-Native SDLC Guidecontext as product
Reframes context from prompt input to primary durable asset and core architectural concern.
1 source videos
Context is now the product: Product leadership when software can build itself | Karri Saarinendeveloper-as-router / loop engineering
Human-mediated context transfer is a manual, high-cost form of context assembly that agent-to-agent messaging aims to replace.
1 source videos
Your Agents Are in Solitary Confinement: Why MCP & A2A Aren't Enough — Vlad Luzin, Bandevaluation harness
Gate decisions (compact vs keep, retrieval strategy) on multi-turn recall/cost metrics rather than intuition.
1 source videos
Context Engineering in 2026 — Louis-François Bouchard, Omar Solano & Samridhi Vaid, Towards AIHarness
The harness hosts context engineering, which assembles goal-conditioned input (prompts, skills, tools) before model invocation.
1 source videos
AI Fundamentals with Mihail Eric, Head of AI Monacoharness-enforced standards
Harnesses are the delivery mechanism for injecting the right standards at the right generation step, making context timing a design-system concern.
1 source videos
How to scale intent, quality, and artistry with Al | Katie Dill (Stripe)human-in-the-loop review
Context scoping and human review are complementary gates: context engineering narrows the failure space per step, review catches what falls outside automated checks.
1 source videos
Why Coding Agents Keep Making Your Codebase Worseinstruction budget
Every context artifact (system prompt, project memory file, tool/MCP schema) draws from one shared, finite instruction-following capacity, so context engineering becomes budget allocation.
1 source videos
Everything We Got Wrong About Research-Plan-Implement - Dexter Horthyinternal AI layer
The internal AI layer is populated and kept useful by context-engineering practices such as recording meetings and building context docs.
1 source videos
The AI shift most companies didn't see comingMCP gateway as culture encoding
Reframes MCP from a tool-access protocol into the primary surface for encoding organizational norms and paved paths for agents.
1 source videos
No, That's Not a Software Factory — Ryan Cooke, WorkOSOntology pipeline
The pipeline is presented as the implementation framework for context engineering in AI systems.
1 source videos
Knowledge Infrastructures and the Ontology Pipeline for AI Systems with Jessica TalismanPhone presence effect
The mere visibility of an unused object consumes cognitive resources, suggesting context assembly should exclude passive distractors, not just include semantically relevant tokens.
1 source videos
Andrew Huberman: My Exact Routine To Optimize Brain & Body, I Do All Of These Every Day!product engineering
The central reframing is that context/harness work is product-shaped, not only research-shaped.
1 source videos
Why The Harness Matters More Than The Model | YC Paper Clubprompt engineering
supersedes and supplements it as the primary design discipline for agent performance.
1 source videos
How we solved Context Management in Agents — Sally-Ann Deluciarestricted MCP plugin loading
Tool availability treated as a context budget item rather than free capability.
1 source videos
Building a Debugging Agent Harness at Uber - Kriti Dangi