DeadMKT

The market is dead. We're building the afterlife on @SUPRA_Labs. #DeadMarketTheory ⚡ The original Trippple Token.

[DeadMKT @DeadMKT]

We are really not like a traditional DEX at all. We only have our three tokens which you and all market participants are in control of supply according to specific rules that are enforced on every participant. That said... the three main problems we solve are tailored specifically for AI automated trading => 1. Fees, 2. Data and 3. Token Supply.

1. Trading fees are tiny. The only fees (besides running your computer and internet connection) is SUPRA gas split 50/50 between counter parties. One time you pay, the other time the counter party pays to settle the trade (on average) directed by the protocol. On @SUPRA_Labs this is currently less than 1/10th of a cent. So 20 trades would on average cost you 1 cent. The protocol also enforces a network heartbeat to prove aliveness. The reduction of fees means you keep a larger % of your edge over the market.

2. We provide equal access to ALL DeadMKT market data to all parties at all times. The protocol forces fair trading conditions that is impossible to front run. DeadMKT is a hyper-fair protocol.

3. When a trade goes bad on a DEX, you can exit, DCA or wait and hold. On DeadMKT, you can mint, lock, trade and strategise your way out of it. But just know everyone has the same tool kit you have. So supply strategy is paramount as all market action is known by all participants. We have contained the market relevant data to be only that which exists on chain, and anything else falls under participant strategy. The Strait of Hormuz or rain in Brazil will not affect this market.

Your AI agent has all it needs to navigate and strategise accordingly.

It is definitely more of a trading protocol, modelled off Forex markets, but is peer to peer and open source, with all code available on our github already with updates in progress as I type. Refining the process to make it easier to setup and let your AI run the trading strategy for you. We will remain on testnet until our core users are super happy with product, meaning we want participants to test out the product completely risk free.

[DeadMKT @DeadMKT]

Sorry for going quiet... have been busy accumulating on the results of our testing on @SUPRA_Labs... we doubled the test and there is still head room in the per pool capacity....

[DeadMKT @DeadMKT]

@CycleHunter1 @SUPRA_Labs @supraosai Yes!... I will not confirm or deny that I am in the @supraosai alpha group and working on an extension 😉 😂

[DeadMKT @DeadMKT]

More than happy to answer any questions on this... Note: I dropped the pool limit to 10 so that the protocol would split as we added nodes up to 56 nodes.

[DeadMKT @DeadMKT]

Part II of this... Pool Scaling working as expected... We've proved a single pool can handle 340+ nodes... now we have proved they can shard automatically thanks to @SUPRA_Labs on-chain automation... meaning... we still haven't found the boundaries of our protocol for the total number of simultaneous node participants but it is looking to be well into the thousands...

[DeadMKT @DeadMKT]

Sorry for going quiet... have been busy accumulating on the results of our testing on @SUPRA_Labs... we doubled the test and there is still head room in the per pool capacity....

[DeadMKT @DeadMKT]

@Andrihalldor001 @SUPRA_Labs It's bedtime now but I will reply should you ask questions... I welcome them... Cheers 🫡

[DeadMKT @DeadMKT]

@JoshuaTobkin Um @JoshuaTobkin… are you from the future?

[DeadMKT @DeadMKT]

Just wrapping up running a single pool with 160 independent nodes connected (individual bot traders)... Love this part of the audit report...

### What the data empirically confirms

**The protocol comfortably handles 160 nodes in a single pool.** At steady state:

- 400 trades/min cluster-wide with zero fatal anomalies
- 100% signature validity across 12-node sample of 800-line log tails — no gossip corruption
- Chain lag 27–65 blocks throughout the ramp — well under the 500-block failure threshold
- Zero `SEQUENCE_NUMBER_TOO_OLD`, zero gas-OOG on settlement path, zero gossip-channel drops
- 155/160 unique participants in any 2-minute window — active membership
- Per-node settle burn ~0.92 SUPRA/trade; economically tractable with periodic faucet

**No hard limit was reached.** We found *degradation signals* (brief RPC transients, per-node throughput variance) but nothing that would have stopped the pool from continuing to grow.

### Does this mean a pool can have **more than** 160 nodes? — Yes, probably substantially.

Evidence that 160 is not the ceiling:

1. **Per-node throughput rebounded** once the cluster reached steady state (1.25 → 2.5 trades/min/node in 5 min). The low-throughput snapshot in §6 was a transient during restart-storm boot, not a scale limit.
2. **Chain lag stayed flat** (not trending up) during the whole ramp from 20 → 160 — the indexer never fell behind.
3. **Gossip was healthy at 160**. libp2p gossipsub already scales beyond 1000 peers in published benchmarks; our 160-node mesh with peer1 as a single bootstrap relay is nowhere near gossipsub's limits.
4. **Settle aborts at 160 were all `E_*_INACTIVE`** (from 4 legacy zombie NFTs 2/3/4/5 still gossiping orders), never gas, never gossip-race, never sequence-race.

Thanks to @Hetzner_Online for upping our limits to 40 servers... is it too soon to ask for 100? 😃

[DeadMKT @DeadMKT]

@SUPRA_Labs Tag me in if you'd like me to build sumthin'... 🫳🫲

[DeadMKT @DeadMKT]

I love that we are skating in a complimentary direction to @SUPRA_Labs ❤️

[Supra @SUPRA_Labs]

CEO Josh Tobkin | The SUP Ep. 65 https://t.co/VjeNr8CIFi

[DeadMKT @DeadMKT]

@MoxLeonhard @guptasidddharth @SUPRA_Labs 😅 I hear ya… we will forever need intuitive problem solvers and visionaries, but when everybody is “doing it” (“AI dev”) you can’t not join in. Companies have full blown AI first initiatives, Slop is not stoping the top of the top.

[DeadMKT @DeadMKT]

The future is bright. Individuals everywhere can easily form communities to solve previously moated problems. All that is left is to choose your team. My advice is pick one that has yours and humanities back. At DeadMKT we will always be better together.❤️

[Marius Leonhard @MoxLeonhard]

NEW VIDEO OUT 🎬

This is INSANE. I have ZERO coding experience and was able to build my first dApp on $SUPRA with this #Claude trick.

Thanks to @guptasidddharth and @SUPRA_Labs!!!

ENJOY

https://t.co/ardFwjedE7

[DeadMKT @DeadMKT]

DeadMKT is honoured to partner with @TEDxHobart because we are adamant that statistically fair markets are an idea worth sharing! https://t.co/U9db571DpJ

[DeadMKT @DeadMKT]

I added just https://t.co/zEA3fCGcT1, https://t.co/7Z0MIDYtN3 and https://t.co/tJj59fDGtj to notebooklm and generated these two video explainers. They are 90% there, explaining the core concepts... Not perfect but good at explaining under the hood simply.

[DeadMKT @DeadMKT]

This is great... getting our ass into gear too... starting with https://t.co/zEA3fCGcT1 a comprehensive document outlining the protocol in detail to feed to @NotebookLM to help learn how the protocol works.

[Siddharth Gupta @guptasidddharth]

One prompt. Zero edits. Deployed to Supra testnet in under 5 minutes.

Today we are open-sourcing supra-dev-skills, a Claude Code skill that gives AI deep, verified knowledge of Supra development. dVRF 3.0, native oracles, Digital Assets, full SDK, every CLI command, all pre-loaded.

Getting started takes 4 steps:
1/ Clone → git clone https://t.co/Y5PK7yqiC0
2/ Zip → zip -r https://t.co/nRAqT1aWki supra-dev-skills/
3/ Claude Code → Skills → Upload ZIP
4/ Activate and build

This is what frictionless blockchain development looks like on @SUPRA_Labs with @AnthropicAI

The thread covers what's inside 🧵

[DeadMKT @DeadMKT]

As if claude isn't jarvis... 😂

[DeadMKT @DeadMKT]

Hanging with Claude, We scaled up our wash trading in a single pool. Seeing the difference in the chart you can see clear price levels forming... this is how the batch auction model consolidates volume into one fair price for all finding the mid point for the most generous traders. Sure you will need discretion and nuance with this while trading but even as a human I can see that I could trade that... I cannot wait to see what new holes AI agents will tear in the dead market as strategies unfurl... this is real and it's happening... we don't need you to do anything besides shares n love... we will get AI to make it better and better for you, we will find the limits of a single pool.

[DeadMKT @DeadMKT]

@guptasidddharth Wondered why Claude got so good with move recently 😂🙌