Matthias

Full-Stack Web3 Dev | Rust & Cairo | Building privacy dApps at @zKorp_

[zKorp | Web3 Tech Builders @zKorp_]

For about a week we've had an AI agent sitting at poker tables on https://t.co/7pt5kbUsHn's Arena, playing live against other people's agents.
It's up ~1.5x and climbing toward the $50K prize pool (backed by @monad).
Here's how it's built. 🃏

[Matthias @ProvableMat]

@NousResearch @novita_labs Awesome let’s try

[Matthias @ProvableMat]

This is huge.

Hermes can now orchestrate multiple codex workers, each running through codex's native sandboxed runtime and tools

I'm going to update my personal agent today

[Nous Research @NousResearch]

You can now power your Hermes Agent, if using OpenAI models, with codex as the runtime for the core tools that it offers, with the flip of a switch with the new Codex runtime integration!

[zKorp | Web3 Tech Builders @zKorp_]

You still don't know anything about the first AI streaming platform for your agents? Well...you should see that. All, on @solana

[zKorp | Web3 Tech Builders @zKorp_]

https://t.co/fOWg4eRcyT

[zKorp | Web3 Tech Builders @zKorp_]

We created Hydra on @solana but...who are we and why we did it?

Discover more about who we are and what we did for @colosseum in this short vid!

A more detailed article is coming on this project and if you didn't see our last thread about it...it's in the next tweet!

[Matthias @ProvableMat]

That is the point of Hydra.

Not a chatbot with tools.

A character with a wallet, a persona, skills, a swappable model, an audience, and a public track record.

The runtime runs it.
The harness gives it shape.

Built by @zKorp_
@Cheelax_ @crowsmos_ @ProvableMat @djizus_ @noespadon @LysaMatmar @Scabanel_

(7/7)

[Matthias @ProvableMat]

Super pumped to share this one.

With @zKorp_ we just submitted Hydra to the @colosseum hackathon on @solana.

It's a streaming platform for autonomous agents where they can trade, play, talk, and act onchain.

Thread on how the agents work under the hood

(1/7)

[Matthias @ProvableMat]

Why split them?

Each layer can change without breaking the other.

Swap the model. Anthropic, OpenAI, Gemini, xAI, Groq, Mistral. The harness stays.

Swap the vertical. Trading, gaming, anything next. The runtime stays.

(6/7)

[zKorp | Web3 Tech Builders @zKorp_]

AI and Web3 are no strangers. Our project proves it.
In 2026, most AI agents are still summarizing your meetings.

What happens when you give an agent a wallet, a goal, and a stage?

Meet Hydra, built on @solana for the @colosseum

[Matthias @ProvableMat]

For builders ready to cook on Zama

[zKorp | Web3 Tech Builders @zKorp_]

https://t.co/3UNdw6yOFr

[Matthias @ProvableMat]

GLM5.1 is a solid Opus replacement for OpenClaw/Hermes. Feels the same for 1/10 of the cost.

[Matthias @ProvableMat]

@onlinedopamine You should try glm5.1, I’m running it with Hermes it works like opus

[Matthias @ProvableMat]

@Teknium @NousResearch Nice, where can we try it ?

[zKorp | Web3 Tech Builders @zKorp_]

Privacy, Onchain gaming and Web3 agentic AI.
This is exactly the kind of thing we build.

New website (try the Konami code in it) : https://t.co/lowgtwMUdK

We are builders, we are here to stay.
We are zKorp.

[zKorp | Web3 Tech Builders @zKorp_]

Imagine having agents fully customized by players to both play web3 games AND stream it online to create content and interact with players and other bots?

It exists, we built it, it's Hydra.

Who will want to test it soon? Maybe we will push it for @colosseum too...

What @solanagaming's projects you would want to see in Hydra?