Anthropic engineer: “80% of our engineers are using self-improving loops. Now everyone is exploring how to build the right Context for agents.” in 3-6 months, we’ll all become context engineers. Context is the most important part of building agents.” In this 40-minute lecture, an Anthropic engineer shows how to... build a self-improving agentic workflow from scratch. It’s worth more than a $500 agentic course. Watch this video, then read the article below on how to become a context architect.show more

80% on self-improving loops and the next frontier is context. context is the part most people skip

yeah, this is true. context is most imporant part if agentic systems

This hits. I’ve been spending more time cleaning context than writing fancy prompts lately

Yeah, this is true. That’s why we need to learn how to manage context correctly. This video will give you the best breakdown.

Appreciate it. Gonna watch it properly later today. Thanks for sharing!

prompt engineer, context engineer, graph engineer, new title every quarter and the job is still choosing what the model gets to see kinda love that the boring part is the one that keeps paying

This course will help you in education, and will definitely be useful for me personally

“Context architect” might sound early now, but I can definitely see this becoming a real skill.

Nice find, Movez. Saved this one.

self-improving loops sound insane next we need agents that improve our groceries

Yeah, but context engineering is a very important part of any agentic system.

loops shift leverage from prompts toward feedback

The best context isn't the most context. It's the right information at exactly the moment the agent needs it.

Builders sleeping on context engineering are about to get left behind

ngl this is exactly what i been building with forge. context engineering is the whole game now fr

Yeah, in my opinion, context engineering is one of the most important parts of agentic systems.

The hardest skill may become organizing knowledge so AI can actually use it well.

Context engineering will probably become the new prompt engineering. Better inputs, memory, and workflows will define better agents.

Bangerr

context engineer is about to flood linkedin bios

prompt engineering just got rebranded

Context engineering is the next title. The hard part starts when multiple agents start writing into the same context and no one owns it

A self-improving loop relies entirely on creating tight feedback mechanisms so the model can evaluate its own intermediate steps. When agents are provided with clear contextual boundaries, error recovery becomes systematic rather than accidental.

Ngữ cảnh là thứ quyết định tất cả luôn.

the best engineers use the best methods, and that’s worth paying attention to if you want to get good at using AI

agents need cleaner windows now

genuinely curious how many of you have actually tried building something with AI — not just using it for quick tasks