Codez

LangChain founder: "Context windows are larger, but they're still not infinite. At some point you need to compact. Context engineering describes everything we've done at LangChain without knowing that term existed." In 40-minute video, LangChain founder reveals why context is the real skill behind agents. compaction + file systems + memory = context engineering for long-horizon agents. Watch the interview, then save the framework below.

@Av1dlive Thanks, Avid ! Happy it’s useful for you

@0xCodez cool share

@gippp69 Context engineering is the most important part.

@0xCodez compaction and managing memory feels underrated, but it’s literally what makes agents practical for anything long running or multi step

Andrew Ng just released a 2-hour course on building agentic skills from scratch with Anthropic: • 00:00 – How to build agent skills with Claude • 22:32 – Claude pre-built skills for AI agents • 41:07 – Agentic skills vs tools, MCP, subagents • 01:06:06 – Skills for long-running agents This 2-hour watch will replace 10 paid courses on building agents, by Anthropic themselves. Watch it today, then read how to build self-improving agentic systems in the article below.

@Granite0x you are welcome mate, happy its useful for you !

@0xCodez timestamps really matter here thanks for adding them started watching from 22:32

@gippp69 Yeah. This is the full guide for building agentic systems skills. Real alpha here

@0xCodez breaking down the exact timestamps for things like pre built skills and mcp subagent design is a total timesaver for deeper dives

@slash1sol top tier read mate. using Minara on daily !

@0xMortyx Yeah, building agentic skills is one of the most important parts of the whole agent architecture.

@0xCodez Agentic skills vs tools its the most important part in my opinion

@0xMovez this is the best course on agentic systems i have ever watched, thanks for the share Movez

Andrew Ng just dropped a 2-hour course on building self-improving agentic systems from scratch: • 02:23 – Introduction to AI agents • 14:35 – Agentic workflows & design patterns • 55:50 – Practical tips for building agents • 01:22:53 – Self-improving agents (evals, loops) • 01:32:42 – Multi-agentic systems This 2-hour watch is worth more than 10 paid agentic courses on the internet. Watch it today, then read how to build self-improving agents with Horizon in the article below.

@0x_rody top tier watch bro ! thanks for the share

Anthropic Head of Engineering: "You prompt agents. The people ahead turned them into infrastructure that works while they sleep." In 16 minutes, three Anthropic leads show the infrastructure your agents are missing, and the exact mistake you make. 1:00 -Claude Managed Agents in production 3:00 - thinner harnesses and adversarial agent pairs 8:20 - the real adoption barriers: security, compliance, evals 9:15 -how to actually measure agent ROI 12:45 - failure modes: hyper independence and sprawl 13:30 - agents as an invisible substrate Watch it, then save the exact config below👇

Andrej Karpathy just dropped a 6-hour course on how to build LLMs from scratch: • 00:00 - Deep dive into LLMs like ChatGPT • 03:31:23 - Building ChatGPT from scratch in live • 05:27:43 - How to use LLMs (Karpathy method) This course will replace a $90K Stanford LLM master’s degree. Start watching today, then read how to become an AI engineer in article below.

@undefinedKi yeah bro. this is an amazing watch, defor worth your book

@0xCodez this is gold. bookmarked

@RetroCoderX Yeah, they just released a Stanford course for free, basically.

@0xCodez Karpathy really said here's a Stanford degree for free

@shmidtqq Yeah mate, that’s crazy what free education can give you nowadays.

@0xCodez bro just made the $90k degree obsolete in 6 hours

@de1lymoon Damn this is crazy system ! Will need to run the same. Thx for the share bro

1 KNOWLEDGE GRAPH + CODE + SOPs + EMPLOYEE DATA = THE $10,000 COMPANY BRAIN MOST TEAMS ARE MISSING Most users store everything in folders, docs and spreadsheets. Then nobody knows how the pieces connect From 0:08 this setup turns the whole company into a visual knowledge graph: • code is linked to projects • projects are linked to agent teams • agent teams are linked to SOPs • SOPs are linked to workflows • workflows are linked to employees • employees are linked to skills, roles and company data The graph shows what normal folders hide: who owns what, which systems depend on each other, what process drives each task, and where the missing context lives. The goal is not to make a pretty map. It is to build a second brain for your company. Bookmark this before building your AI company brain

@shmidtqq Andrej Karpathy is real legend ! Amazing watch. Thanks for the share bro

Andrej Karpathy just dropped a 1-hour breakdown of what AI work looks like in 2026: • 01:00 - Why he barely writes code manually anymore • 09:20 - His “Dobby” agent finds smart-home devices and builds a control dashboard • 17:10 - AutoResearch: agents testing research ideas without humans slowing the loop • 29:20 - Why models can write CUDA but still fail at jokes • 44:46 - Why open source matters when frontier AI gets centralized This is not a podcast about better prompting. It is a podcast about the job moving up one layer. The human stops being the person typing code and becomes the person designing loops, giving context, checking outputs and deciding when the agent is wrong. That is exactly why the article matters. Prompting gets you an answer. Loops get you work done while you are not touching the keyboard.

@gippp69 Amazing watch bro ! Griffin is a real OG

HE STARTED TRADING FROM A HARVARD DORM ROOM WITH A SATELLITE DISH BOLTED TO THE ROOF. TODAY HE IS WORTH AROUND $45 BILLION. Ken Griffin built Citadel into a hedge fund managing roughly $69 billion, starting from a college room wired for live market data. Citadel Securities became one of the biggest market makers in the world, handling a massive share of US retail stock trades every day. in his strongest years running the fund, his personal earnings reportedly landed between $1.5 billion and $4 billion a year. the entire empire traces back to a student who installed a satellite dish because his dorm room did not have access to a proper trading terminal. bookmark this before the next “you need connections to start” excuse.

@gippp69 yeah, live building right in course

@0xCodez does the course include hands-on code or just theory for the live build part?