Greg Johns

AI developer building automation tools for real estate. MCP servers + AI agents. Personal liberty, sound money. Orlando.

[Greg Johns @gjohnsx]

@misscate12 @peptidepirate where are you injecting

[Coinbase Developer Platform🛡️ @CoinbaseDev]

🚀 A BIG week for x402 and payments! Here’s the latest:

Updates:
→ Coinbase for Agents is live. Give your agent its own account to trade, manage your portfolio, and run autonomously under your guardrails.

→ @MurrLincoln gave a demo on the new, simpler sign-in for the Coinbase CLI and MCP using Claude.

→ Our developer wallets are now available via a Swift SDK, so you can build self-custodial wallets into native iOS apps.

x402:
→ @travalacom introduced their Travel MCP: featuring gasless USDC transactions on @base via x402.

→ @Must_be_Ash gave a demo on booking travel with Claude connected to Base and Travala’s MCP.

→ @yugacohler and @programmer talked about agents and x402 at @ethconf.

→ @brian_armstrong shared: 160M+ agentic payments processed via x402 and ~$1T in annual stablecoin movement at Coinbase.

Payments:
→ Our Head of Infrastructure Products, @alecglovett, broke down everything Coinbase has built for stablecoin payments.

→ Cross-border stablecoin payouts are live for @masspay_io, powered by Coinbase.

Coming up:
→ x402 payments with user's retail USDC balances arrive in Coinbase for Agents in the coming weeks, stay tuned for updates.

→ System Update on June 16. The future of payments is built on Coinbase.

[OpenAI Developers @OpenAIDevs]

Introducing developer mode for browser use in Chrome and the Codex in-app browser.

Codex can use the Chrome DevTools Protocol (CDP) to debug browser issues by profiling JavaScript performance and inspecting console output, network traffic, and page state.

[Mnimiy @Mnilax]

Jenny Wen, head of design for Claude at Anthropic:

"the design process is dead."

what replaces it runs on machine-readable design systems

In a 1-hour conversation, Jenny Wen breaks down why the discovery -> mock -> iterate loop is dying and what designers do instead.

design system -> Claude Code -> every screen fixed at once, not one at a time

i rebuilt the same 10-screen dashboard twice:

~6 hours of restyling by hand vs ~40 minutes when the fix lived in the system.

how to build that system and hand it to Claude Code is in the article below.

[Codez @0xCodez]

Anthropic Managed Agents team:

"Fable 5 is our best model for running self-improving agent systems.

Add /loops, dynamic workflows, dreaming and you are unstoppable"

in 13-minutes, Anthropic team shows how to build self-improving agent systems with Fable 5 from scratch.

Worth more than a $500 agent building course.

Live from the last Anthropic stage in Japan. Unpublished.

[Greg Johns @gjohnsx]

@Scolmj @Erik_Hultstrom @manutdaremuck @MarcBhafc This game was in Orlando

[Greg Johns @gjohnsx]

@Molson_Hart Amazing product. Just got ours and we’ve all been playing with them for 2 hours.

[Anthony Kroeger @kr0der]

btw this is one of the best codex plugins if you're doing web dev

it gets the agent to perform testing every time that it does frontend work, including starting dev server -> in-app browser testing -> taking screenshots for you to review

it even tests mobile viewports too

[Yohei @yoheinakajima]

slowly we're all realizing that tools should be called from code, not from within the llm api

[Greg Johns @gjohnsx]

@metropol same, all the time. hardly ever able to connect. no solutions yet

[Jonathan Goodman 🇨🇦 @itscoachgoodman]

You only get:
a few years of them crawling into your bed
a few years of them running to the door when you come home
a few years of them asking for one more chapter
a few years of them needing your help
so read the book one more time,
push the mower together,
lie in bed until they fall asleep,
stay for the slow morning
because the days of being needed this completely
do not last forever.

[OpenAI Developers @OpenAIDevs]

Building voice applications with GPT-Realtime-2?

Our new prompting guide covers how to tune reasoning effort, use preambles, design tool behavior, handle unclear audio, capture exact entities, and maintain state in longer sessions.

https://t.co/9zfdhIX4Vq

[jason @jxnlco]

新しいリアルタイム翻訳モデルを発表できることをうれしく思います。ぜひ本日よりAPIでお試しください。

[Greg Johns @gjohnsx]

@g_snso @nassy0708 Wish I had known that yesterday

[Greg Brockman @gdb]

build your own agents with codex app-server

[Romain Huet @romainhuet]

You can now work on @supabase apps directly in Codex!

[Greg Johns @gjohnsx]

@Champdebloom @codetaur @RepoPrompt yeah trying to figure out how to just get codex app to use repoprompt mcp along with with gpt 5.5 pro with browser use

[Greg Johns @gjohnsx]

@rtheoryxyz @maxescu nft comeback time

[Y Combinator @ycombinator]

Company Brain
@t_blom

Every company has critical know-how scattered across people's heads, old Slack threads, support tickets, and databases, and AI agents can't operate like that.

We think every company in the world is going to need a new primitive: a living map of how the company works that turns its own artifacts into an executable skills file for AI.

[Romain Huet @romainhuet]

The Realtime API is fast.

This demo by @jxnlco is running at real speed! No computer use, just realtime speech + function calling.

The model is listening, deciding when to call functions to update state, and yes, it doesn’t always have to talk back!