@0xanand
ANAND VASHISHTHA
@0xanand
A dev with too much entropy.
A dev with too much entropy.
Ghaziabad, India
Devfolio stats
Devfolio stats
5
projects
5
1
prize
1
23
hackathons
23
1
Hackathons org.
1
GitHub
GitHub
1,010
contributions in the last year
Oct
S
M
T
W
T
F
S
Sep
19
stars earned
71
repositories
67
followers
KaggleIngest
Extract context from Kaggle competitions and datasets for AI coding assistants. Fetches top notebooks, metadata, and dataset schemas into LLM-ready formats.
Python
0Stars
0Forks
coremap
Pack a repo into one AI-friendly file — or keep only the lines that matter for one task, with a verification receipt.
TypeScript
5Stars
0Forks
0gskills
HTML
0Stars
0Forks
adreceipt
TypeScript
0Stars
2Forks
Medium Posts
Medium Posts
AI Agents Don’t Need Crypto. They Need Authority.
medium.com
The real convergence of AI and crypto is not about giving machines wallets. It is about deciding what machines are allowed to do. For most of crypto’s history, the idea of an autonomous machine economy arrived too early. Projects imagined AI services discovering one another, exchanging money, building reputation and collaborating through decentralized markets years before modern AI agents existed. The vision was not necessarily wrong. The actor was missing. Early “agents” were mostly narrow models, bots ...
WebMCP: The Web Was Built for Humans. AI Agents Need Something Else
medium.com
The web browser has always been designed for humans. We read text, recognize buttons, understand menus, fill forms, and decide where to click. AI agents currently have to do much of the same thing artificially: inspect the DOM, analyze screenshots, locate elements, click buttons, wait for page changes, and hope the interface has not changed. That works — but it is a fragile abstraction. A checkout button moving should not break an AI agent’s ability to buy something. A redesigned form should not force an...
Perplexity Tried Advertising. Then Trust Became the Product.
medium.com
The AI search company’s retreat from sponsored placements reveals why advertising becomes more complicated when software gives one synthesized answer. In November 2024, Perplexity began experimenting with advertising in the United States. Its format was relatively restrained: sponsored follow-up questions beneath a response and paid media beside it. Perplexity said the advertisements would be clearly labelled, that sponsors would not write or edit generated answers, and that advertising would not change ...