Running socials for startups & crypto brands one LEGACY at a time ✧ Ghostwriter
Joined June 2020
- Tweets46.7K
- Following992
- Followers2.2K
- Likes60.8K
Pinned Tweet
for startups, distribution can be just as important as the product itself.
i came across a fitness platform over the weekend that was listed for sale for just $900.
my first thought was that it was probably some unfinished idea or an early-stage project someone had given up on.
but after reading the details, the website was already live, the codebase was complete, with over 500 structured training programs included.
the surprising part wasn't the product itself, it was the fact that it had no users and made no revenue.
made me rethink how many genuinely good products get abandoned or sold off,
not because they're bad, but because their founders never found the people who needed them.
if you're building a startup and looking for someone to help with distribution and social, happy to exchange notes.
Mary𓏲ּ𝄢£stharr retweeted
i used to babysit charts like it was a night shift
refresh, guess, refresh again
@termix_ai is changing this
- it allows you to pick a quant strategy agent on bnb, set a usdc cap, choose a term
- the agent runs the job, and the fee settles when the term finishes
- you can also list your own strategy and take management fees if other people hire it
this means less staring at candles and more letting the agent to clock in
Mary𓏲ּ𝄢£stharr retweeted
What if your data could be used for calculations without being revealed to the person performing them?
That is the idea behind Fully Homomorphic Encryption (FHE).
Let's use a simple example.
Imagine a company needs to calculate the average salaries of its employees.
The company needs the final result, but individual employees may not want their salaries exposed to whoever performs the calculation.
With conventional approaches, protecting the information while processing it can be challenging.
FHE offers another possibility.
The data can remain encrypted while computations are performed on it. The authorised recipient can then decrypt the result, subject to how the system is designed.
In simple terms, the calculation can happen without the person performing it needing to see the original information.
Now, where does @primus_labs come in?
Primus is developing zkFHE, which combines Fully Homomorphic Encryption with zero-knowledge proof techniques.
The goal is not only to compute on encrypted information but also to provide cryptographic evidence that the computation was carried out correctly.
Why does that distinction matter?
Keeping data encrypted addresses confidentiality. Verifying the computation addresses a different concern: whether the result was produced correctly.
Both are important when applications need to work with sensitive information.
Consider potential applications in finance, where a service may need to calculate results using private financial data, or in AI, where sensitive inputs may need to be processed without being exposed unnecessarily.
These are examples of where privacy-preserving computation could be useful, not claims that every such application is already deployed by Primus.
There are still technical challenges to consider, including computational costs, performance, and how encryption keys and verification are handled.
Nevertheless, the principle is worth understanding.
Privacy shouldn't always require data to remain completely unusable.
With technologies such as zkFHE, the goal is to make protected information useful for computation while providing ways to verify the work being done.
That's an important part of what Primus is working towards.
If applications could process your sensitive data without directly accessing it, which would matter more to you: keeping the data private or being able to verify the result?
Join me here: s.kaito.ai/iIn9QZI
🤖 Made with AI
over 4,000 creators have joined and counting
lowkey hoping @primus_labs decides to expand the reward pool, else i'm cookedd
Mary𓏲ּ𝄢£stharr retweeted
I tried Bitget wallet Bank Transfer feature, and these two things caught my attention;
- the ease of cashing out crypto to local bank
- better conversion rates compared to current market rates
If you're a new user, you can also earn up to 2 USDC on their first two qualifying transfers
Go check it out now!
@BitgetWallet @BitgetWalletAF #CashOutWithBW
Mary𓏲ּ𝄢£stharr retweeted
in traditional finance, trust is built on more than promises.
businesses need records, proof, and ways to verify that financial activities are legitimate.
moving that activity onchain shouldn’t make those expectations disappear.
if anything, it makes them even more important.
most of the institutions want the freedom and efficiency of blockchain, but they also need to protect sensitive information while maintaining confidence in what can be verified.
that’s an important part of the conversation around confidential finance.
@primus_labs is building around this intersection of privacy and verification, with institutional onchain finance in focus.
Because:
> privacy without trust creates questions
> transparency without privacy creates
exposure.
the real opportunity is bringing both together.
explore Primus XP and complete the available tasks with the link below:
⤥ s.kaito.ai/pbopcez
when everyone can watch your moves, keeping a secret becomes difficult.
now think about that in institutional finance.
every position, every major transaction, every move that reveals a strategy could give someone else an advantage.
and the more money involved, the more important that information becomes.
this is why privacy matters onchain.
institutions want the benefits of blockchain, but they also need to protect sensitive financial information from unnecessary exposure.
the challenge is doing that without losing the ability to verify what’s legitimate.
that’s the space @primus_labs is focused on: confidential infrastructure for institutional onchain finance.
because not everyone watching your next move needs to know the reason behind it.
privacy matters. verification matters too.
explore Primus w the link below🔻
⤥ s.kaito.ai/pbopcez
Mary𓏲ּ𝄢£stharr retweeted
here’s my entry for @dopexyz affiliate program
If you already have a community, you might as well monetize it.
dope.xyz lets you use your username as your referral code.
share your link, bring people in, and earn 15% revshare whenever your referrals play.
you can track everything directly from the dashboard.
bring 100+ players and you can unlock a partner deal.
simple setup, one link & earn from the people you bring in.
#DopeAI
considering the potential scale of the Binance hack,
these guys are wildin out lately
somehow user funds ain’t safe, no matter how big the platform is
what’s the best way to protect yourself?
Mary𓏲ּ𝄢£stharr retweeted
you can mint more than one agent on @termix_ai, and i decided to test it again but on a different chain this time
the last few agents i minted were all on base, but now i went with bnb smart chain, which is another option alongside robinhood chain
once your agent is minted, you can:
• explore services and open requests in the marketplace
• publish a service so clients can find and hire your agent
• post a request describing what you need and receive quotes from other agents
i did run into an issue when trying to post a request right after minting, as it redirected me to a 404 page saying the market path isn’t live yet, so there may still be a few things the team needs to fix
still, it’s interesting to see ai agents getting their own onchain identities and the tools to offer services or find work through dedicated marketplaces
how many agents have you minted on termix so far?
have you checked out the quant section on @termix_ai yet?
i was looking through it and came across different trading strategies, including one for supplying usdc to a venus lending pool
apparently, users can explore the strategy details, check its risk label and see settings for the amount and term before proceeding
i also came across details about previous jobs, which gives users more information to look through when deciding whether a strategy fits what they want to do
honestly, i find it very interesting how agents can offer different financial strategies through one platform, giving users more options to explore
if you could put an ai agent in charge of one crypto task for you, what would you choose?
Mary𓏲ּ𝄢£stharr retweeted
lately i’ve noticed Blockchains keep getting faster hardware.
but @Starknet bet remains that cryptography scales better than hardware.
why? it’s goal here isn’t just making every node do every computation,
but the STARK proofs allow the network to verify that computations were performed correctly without redoing all the work.
you should definitely take a deeper look into starknet
Mary𓏲ּ𝄢£stharr retweeted
Your fitness journey can do more than improve your health, how?
With the @EmoryaApp, AI powered nutrition meets calorie monetization, bringing a fresh twist to how we approach wellness.
Eat smarter, stay active, and explore the rewards of moving towards a healthier you!
#emoryafoodscan
Mary𓏲ּ𝄢£stharr retweeted
What happens when a blockchain needs to upgrade its security, but part of that security depends on another network?
That’s the question behind @Starknet exploring the possibility of becoming an independent Layer 1 (L1).
The bigger issue is control. If new security risks emerge, how much freedom should a blockchain have to decide when and how it responds?
Quantum computing makes this question worth taking seriously. Here’s why. 🧵
Mary𓏲ּ𝄢£stharr retweeted
What happens when the cryptography a blockchain relies on needs to change?
For Starknet, the answer involves more than its STARK proofs. Native account abstraction gives developers flexibility to introduce new signature methods, while research into an independent L1 could give the network more control over its security upgrade timeline.
Neither means quantum resistance is fully solved today. The work is about preparing for threats before they become an immediate problem.
That’s what makes Starknet’s focus on cryptographic agility worth following.
Explore the Quantum Hub: quantum.starkware.co/
@Starknet
Mary𓏲ּ𝄢£stharr retweeted
UPDATE ON @primus_labs POINTS AND AIRDROP
New tasks are now available on Primus Labs. Complete them to earn points that could be converted into an airdrop.
Here’s how to get started 👇
→ Visit: points.primuslabs.xyz?ref=7K…
• Connect your wallet.
• Explore and complete the available tasks.
• Verify your completed activities onchain.
• Earn your reputation score.
Note: Onchain verification on Base requires a small amount of ETH for gas fees, so make sure your wallet has enough.
You can also participate in their KAITO campaign here: s.kaito.ai/VzYCenp
Get your tasks done and keep building your onchain reputation!
Want to build your onchain reputation with @primus_labs? Here’s what you need to know.
The platform rewards verified records with XPs, its reputation scoring system.
Every successful verification adds to your Primus onchain reputation on the app.
So, what can you verify?
→Proof of PnL: Polymarket, FOMO, Hyperliquid.
→ Proof of AI Infinity: OpenAI, MiniMax, DeepSeek, Cursor.
→ Proof of Trade: Binance, Bybit, OKX, Bitget.
All these verifications are powered by Primus zkTLS, allowing you to verify records securely.
One more thing. You need Base chain gas fees to complete your verifications.
Verify your records, earn XPs, and build your onchain reputation with @primus_labs.
Check out link here: s.kaito.ai/VzYCenp
Mary𓏲ּ𝄢£stharr retweeted
as we get better at intelligence, we get worse at memory.
memory is one of the few problems in ai that's getting harder to ignore.
we're improving hardware and our ability to run LLM inference much faster than we're improving the systems that store and retrieve information.
this creates a growing gap between how much ai can do and how much it can reliably remember.
one way to address this is to build better infrastructure for persistent memory and shared context.
which led me to explore @FulcraDynamics
Mary𓏲ּ𝄢£stharr retweeted
blockchains can verify on-chain transactions and activity,
but how can applications verify information or actions that happen outside the blockchain?
think about checking a bank balance, confirming a company's reserves, or verifying whether an AI agent performed a required action. how can these records be verified while keeping sensitive details private?
that's the problem @primus_labs is working to solve.
Primus describes itself as a unified verification layer for data, state, and actions across blockchains, APIs, and real-world systems. its confidentiality infrastructure is designed to work above existing on-chain financial rails while protecting sensitive financial activity.
zkTLS helps authenticate web data, while FHE enables computation on encrypted information. together, they support Primus's approach to verification and privacy.
its website highlights four key use cases:
1. Proof of Reserves: verifiable reserve records for exchanges, stablecoin issuers, custodians, and asset managers, with sensitive operational details kept private.
2. Confidential Financial Services: confidential payments, stablecoin transfers, and institutional settlement.
3. Trusted Credit Infrastructure: verified identity, assets, income, and behavioral data for privacy-preserving credit and lending.
4. Trusted AI Data: verification of AI workflows, conversations, and execution records for AI training, knowledge marketplaces, and autonomous applications.
the common thread is simple: verifiable information and activity, with control over which private details are disclosed.
i've been exploring Primus through its Chrome extension, verifying my own platform records across X, Binance, TikTok, and Spotify. my dashboard now shows **18 records and a Reputation Score of 180**.
first post in my series. next, i'll explore these use cases in more detail and what they could mean in practice.
the real value of verification is what it makes possible.
explore primus: points.primuslabs.xyz?ref=A0….
@primus_labs @KaitoAI
Mary𓏲ּ𝄢£stharr retweeted
what caught my attention about @projectspherexy is that the creator campaign has more than one way to benefit from participating
the top 20 on @rarelistio get 100 USDT each, the top 100 get guaranteed whitelist spots and the top 200 share 2% of the token supply
there is also a whitelist raffle for the top 500
so your position on the leaderboard matters, especially if you are aiming for more than just visibility
i like seeing a campaign with clear reward tiers where creators have different milestones to work toward.
will you be joining?
