BREAKING: Sui is opening a public mainnet experiment on July 4.
Anyone will be able to log in with email and spam fully sponsored transactions through “tunnels,” with the goal of hitting 1,000,000 TPS.
Why keep burning tokens to search and parse your own data yourself by wasting your token?
That’s exactly what we provide: your knowledge, expertise, and condensed chat histories from sessions can become valuable assets for others.
They save people significant time and money while delivering real value.
Helping others do the job correctly and saving time with their agents — that’s what our knowledge layer is for.
It's been exactly a month since Sui suffered one of its biggest outages. The core issue is that Sui brought a breaking change that introduced "address balance" to Sui — a concept fundamentally at odds with its core object-model design.
You can find more context in my summarized article on @ZingonSui, but from a technical standpoint, it forced significant updates to our balance retrieval logic for devs.
With the zing MCP tool integration, you can easily inject that summarized understanding into any agent you're working with, in any workspace.
Your agent doesn't need to scrape web pages, parse messy HTML, and re-interpret scattered references on its own — it can start from a distilled, accurate summary instead.
Saves tokens. Avoids hallucinations. Gets straight to the correct work
Your AI agent shouldn't just guess. It should query real knowledge.
MCP (Model Context Protocol) is how your AI tools pull information from external sources. It's what lets your agent search and retrieve actual content — not hallucinate.
Using your AI agent tool, you can query Zing's knowledge base through MCP. Install the CLI, run the MCP server, and your agent gets access to searchable, retrievable content from any knowledge base on the platform.
Install:
`cargo install --git https://t.co/HuKYBcFN29 zing-cli`
Make sure your Sui wallet has USDC. Run `zing mcp serve`.
Now your agent can search and retrieve content from any knowledge base on Zing. Pays per query. You see exactly what you paid for.
Docs: https://t.co/i4kxzF5Dwf
@EmanAbio Nice article! The Sui ecosystem never stops
You can also publish it here: https://t.co/jbxThhD88N
Let others subscribe and follow Sui’s latest news
If you’re a developer, on-chain researcher, or crypto-financial writer who’s tired of deep work disappearing after launch week — this is built for you.
Open beta. Google login (no seed phrase). First article in minutes.
Try it: https://t.co/7uxvPMoHRs
Most people writing deep technical content already know how it ends:
a week of attention, then silence.
That’s not a bug. It’s the business model. Your work generates the value. Platforms keep the upside
For developers: there’s also an MCP server.
zing mcp serve connects your knowledge base to Claude Desktop, Cursor, VS Code, and any MCP-compatible client. Agents get clean access via zing by paying USDC
Setup docs: https://t.co/Kri0t3gNnW
Today we get into what's actually holding AI agents back: memory 🦭
Walrus Memory keeps an agent's context portable across apps and sessions, controlled by you and verifiable by other agents. We'll cover how it works and who's already building on it.
4 p.m. EDT / 8 p.m. UTC.
See you there!
The whole point is data accessibility
While Web2 companies have long profited from gatekeeping data, the emergence of the AI era demands a paradigm shift. Access will no longer be controlled by platforms
Welcome to the API-first world, a future should be defined by direct interaction without any middleman
Facebook, X, TikTok, and every other social platform makes money by targeting ads on data users generate on the platform.
So if users control that data instead, can you still build a business that runs on ads?
George's answer is yes, plus the new paths @WalrusProtocol opens up on top.