
Full Stack Web Developer || Mobile App Dev. || Football lover #halaMadrid || Web3 || Crypto
{{ inside curly brackets }}
Joined September 2011
- Tweets138K
- Following46.7K
- Followers116K
- Likes64.1K
ok so the new Proof of Creation tasks on @primus_labs are up under Agentic Ara. Hailuo, CapCut and Dreamina each have a “verify your video creation count” task.
Creator stats are normally just screenshots but here the zkTLS check proves the number without exposing the account, which is the part that actually matters, that’s way better than screenshots.
Only thing here is you need the Primus extension installed first, otherwise the verification won’t run.
I’ve only used CapCut so far so I’m starting there. also Nomad Life, PnL, AI Affinity and Trade are still on the page, and they say new tasks drop weekly.
As a dev, the part I want to dig into next is the Page Core SDK. getting zkTLS into a browser app with WASM and a Vite config is a lot more doable than I expected.
prove the activity, skip the oversharing
I’ve been thinking about how many times I’ve had to prove something online by just handing over way too much.
Applying for a job? send the full CV and every document, proving I own an account? share a screenshot with everything visible and all required documents.
Most of the time, the other side only needs to know one small detail, but we end up giving them the whole picture anyway, that’s the part I kept coming back to while reading about @primus_labs.
With zkTLS, the idea is simple. Instead of showing the raw data, you prove a fact about it. Like “this account is mine”, “this activity really happened”, or “this number is above a certain value”. The other side gets a proof that it’s real and came from the actual source, and the private details stay with you.
What makes it feel practical to me is that it’s not just theory. On the Primus Points dashboard you can already verify things like your GitHub contributions or X account metrics. That’s the same idea at a small scale: prove it’s real without handing over your login or exposing everything else.
And it gets bigger from there. Proof of reserves, verifiable agent activity, confidential finance. It’s all the same question: can I trust this, without seeing everything behind it?
For anyone building, this matters a lot. Apps that depend on real-world data have always had to choose between “just trust the source” or “collect way too much user information.” A proof-based approach gives a third option.
I’m a developer, so I tend to judge projects by whether I could actually build with them. The fact that Primus has an SDK and keeps shipping code on GitHub makes it easier for me to take seriously.
Still early, and I’m sure there’s a lot left to figure out. But proving what’s true while keeping the rest private feels like something the internet should have had from the start.
@primus_labs
Join Primus, explore and stack rewards s.kaito.ai/uNHq2c9
Replying to @CryptoSense_2 @rntprotocol
Love how this shows yields are variable based on utilization, while highlighting smart contract, collateral liquidation and capital loss risks with no guarantees from the underlying technology
43
Replying to @BarabiloT @termix_ai
Their Persistent collateral pools that slash malicious actors really stands out
10
Replying to @dezydank @rntprotocol
RNT's focuses on bridging traditional property markets with on-chain finance through mechanisms like Ember Protocol integration.
4
Almost every industry runs on data. Finance, AI, healthcare, you name it. The catch is that the more data gets shared, the more private stuff gets exposed.
What we actually want is for systems to process information, check that it’s real and give useful results, without anyone seeing the private details.
That’s where privacy-preserving computation earns its place. FHE, MPC and zero-knowledge proofs are opening up ways to work with data without leaking it.
This is the problem @primus_labs is going after. Their idea is to pair verifiable data with confidential computation, so you can use, verify and process information without exposing it.
The two core pieces
- zkTLS lets you prove a claim is true without showing the data behind it.
- FHE lets you compute on data while it stays encrypted.
The goal isn’t to hide everything, it’s to prove only what needs to be trusted and keep the rest private.
Privacy shouldn’t be bolted on after a leak. It should be there from day one, especially as more finance moves onchain.
Why their implementation of AutoHoG caught my eye;
Private transactions on FHE used to be painfully heavy, AutoHoG designs compound gates from standard Verilog circuits and merges subcircuits into a single bootstrapping, which cuts a lot of the work. Reported numbers on 8 consumer RTX 5090 GPUs:
→ ~90ms per full 64-bit confidential transfer
→ 210+ transfers per second
→ No exotic hardware
→ Identical results across nodes
Fast and private, which is rarely both.
To be fair, a benchmark isn’t production, and security and integration still have to hold up in the real world. But the direction makes sense to me.
Replying to @recepdemir097 @termix_ai
Securing payments through onchain escrow is very necessary to build trust and be sure agents deliver
6



