Eva Liu

Jewelry & Accessory Designer; Parsons School of Design Alumni; Instagram @petite_evaliu

[Anatoli Kopadze @AnatoliKopadze]

This is the best 2 hours on graph engineering ever recorded, Andrew Ng breaking down how to build agentic knowledge graphs from scratch:

9:14 - your first working agent
33:11 - how loops actually work
1:02:46 - turning loops into working graphs
1:30:15 - agents that improve their own code
1:49:05 - a system that runs the whole thing for you

I've seen $500 courses that teach less than this video alone.

Watch it, then take it further with my step-by-step guide on graph engineering below.

[Ryven @imryven]

karpathy's second brain runs on plain markdown. no vector database. no embeddings. no $30 memory app. just a folder that never forgets.

here is the exact architecture.

most people use Claude the same way every day. open a new tab. explain who you are. get an answer. close the tab. tomorrow it remembers nothing. you are still the one rebuilding from scratch.

this system removes that job from the human entirely.

three folders. one file. three prompts.

raw/ captures everything. drop it in and never touch it again. immutable ground truth.

wiki/ is where Claude converts everything in raw into structured, linked knowledge. the human reads it. the model writes it.

output/ is where finished work lands. reports, posts, documents. built from the wiki, not from memory.

CLAUDE.md sits at the center. not a prompt. a complete picture of who you are. Claude reads it before every session. automatically. you never explain yourself again.

three prompts run the whole system.

ingest: drop a source, Claude extracts it into the wiki and links it to everything already there.

query: ask anything and it synthesizes across every source since day one.

morning briefing: at 7am Claude walks the vault, flags what's stale, checks for contradictions, sends three lines.

you open your laptop and pick up where the thinking already left off.

karpathy runs it on his own reading. 100 articles. 400,000 words. cross-linked and maintained without him touching a single file.

your vault has 3,000 notes nobody will ever reopen. his read all of themselves by breakfast.

one month in, context stops disappearing. three months in, the vault surfaces ideas you forgot you had. six months in, you never start from zero again.

every app promised a second brain. this is the first one that actually compounds.

full guide in the article below. bookmark this.

[Fran Pradas | Ingeniero de IA @franpradasAI]

🤯 Pídele a CLAUDE o CODEX que te de toda la ARQUITECTURA de tu app en una sola página HTML y un archivo JSON.

Tu código ahora se explica solo:

El HTML te permite visualizarlo.

El JSON es para que otro agente lo entienda.

Te dejo el prompt que usé abajo 👇

[Khairallah AL-Awady @eng_khairallah1]

this is f*cking gold

Andrej Karpathy joined Anthropic five weeks ago.

Two Anthropic seniors just made Karpathy's loop 1000x better with "Graph Engineering"

the agentic systems got 1000x better the moment you wired agents into a graph

I dropped it into my setup. The very first response was different.

Not slightly different. Completely different.

Claude stopped giving generic answers and started working exactly the way I think.

Bookmark it before it gets lost in your feed.

Read it now, then check the article below.

[rvaniaaa @rvaniaaaa]

someone built an AI agent that learns new skills on its own. the only human step is the final approval.

8 agents. one loop. runs without you.

every AI agent stops improving the moment you stop improving it. new workflows appear on GitHub every day. better patterns get published. smarter architectures get shared. none of it reaches your agent unless a human finds it, reads it, understands it, and installs it manually.

that is the bottleneck this system removes.

eight agents. each one has exactly one job.

scout finds new repositories. filter removes the noise before any model touches it. reader builds context from docs first, never from source code. docs explain intent. code explains implementation.

extractor pulls the reusable workflow. score validates it against objective criteria. if it fails, it dies here. generator packages it into a complete skill with examples, commands, and tests.

reviewer asks one question: would an experienced engineer install this without editing it? if no, it never moves forward.

publisher opens the pull request. automation proposes. humans approve. nothing merges automatically.

the result compounds. stronger skills lead to better discoveries. better discoveries create stronger skills.

every new capability your agent gets today required a human to find it.

this pipeline removes that constraint permanently.

the complete architecture is in the image.

[Anatoli Kopadze @AnatoliKopadze]

A Google Cloud engineer just showed how to build a full app with Claude from scratch.

He spent 26 minutes live on stage doing what most teams take weeks to do.

Worth more than any $500 vibe-coding course, no team, no setup, just Claude and a goal.

Learn what Claude can actually do and you'll ship alone what others hire a team to build.

Watch it, then read the guide below on the Claude features 99% of users never find.

[Eva Liu @petite_evaliu]

The model is the intelligence The harness is the machinery that makes that intelligence useful

[rari @0xwhrrari]

https://t.co/9qtKcgOtPH

[Eva Liu @petite_evaliu]

A LOOP is：
DISCOVER → work out what needs doing
PLAN → decide how to do it
EXECUTE → do the work
VERIFY → check it against the goal
ITERATE → not there yet? feed the result back in and repeat

[Anatoli Kopadze @AnatoliKopadze]

https://t.co/eq0oL4fIaR

[Eva Liu @petite_evaliu]

All in Context

[Ryven @imryven]

https://t.co/tGS9Y7blKE

[Ryven @imryven]

https://t.co/tGS9Y7blKE

[Eva Liu @petite_evaliu]

Goal（目标）→ Constraints（约束）→ Deliverables（交付物）→ Acceptance Criteria（验收标准）

[Eva Liu @petite_evaliu]

https://t.co/9cTcyIGC19

Anyone here？

#NFTNYC2023 #NFT #web3

[Eva Liu @petite_evaliu]

@ZR7Goat I am here🙋！

[Eva Liu @petite_evaliu]

@Summerkiwi0612 我之前也在苏富比实习过！

[Eva Liu @petite_evaliu]

Hello Twitter, hello vogue #vogue #annawintour #fashion #pharrellwilliams #gigihadid #stellamccartney