AgentDNA

256 posts

AgentDNA banner
AgentDNA

AgentDNA

@Agent_DNA

Redefining digital trust. AgentDNA encodes authenticity, governance, and context into every intelligent entity

Singapore Katılım Kasım 2025
6 Takip Edilen20 Takipçiler
Sabitlenmiş Tweet
AgentDNA
AgentDNA@Agent_DNA·
AgentDNA is live now! The first unified trust layer for multi agent systems, built on ultra scalable DAG of @RubixChain. pypi.org/project/agent-… Why AgentDNA? 1. Auth: Current static auth systems built for humans/static APIs dont work for Agents. AgentDNA introduces dynamic DIDs
English
2
7
6
646
AgentDNA retweetledi
srini ravi
srini ravi@rock2905·
@rohan_2502 For cross organisation and ecosystem adjacencies involving agentic reality enterprises will require a new unified,verifiable trust layer which is independent of models,hyper scalers and workflows. @Agent_DNA is the new Compliance ready and ROI friendly accountability guardrail.
English
0
2
2
42
AgentDNA retweetledi
KC
KC@KCRubix·
The incident in which @cursor_ai and @Railway agent deleted data without authorization is yet another proof that agentic systems without immutable chain of auth and logs are highly risky. Repeat, Agentic security/governance is SO DIFFERENT from current MFA systems. Build first principles.
KC@KCRubix

If your agentic systems allow agents to alter/delete their logs or auth chains, then there is little security or governance. Immutability/non repudiation is vital in agentic security. Without immutability, rest of the zero trust mechanisms become ineffective. @Agent_DNA

English
0
3
4
114
AgentDNA retweetledi
KC
KC@KCRubix·
Memory is 40% of AI infra cost. Longer token context makes models powerful, but accentuates memory constraints. Efficient graph based context traversal and retrieval using @RubixChain graph will save billions of costs and also scale agents databricks.com/blog/memory-sc…
English
2
3
4
504
AgentDNA retweetledi
KC
KC@KCRubix·
If your agentic systems allow agents to alter/delete their logs or auth chains, then there is little security or governance. Immutability/non repudiation is vital in agentic security. Without immutability, rest of the zero trust mechanisms become ineffective. @Agent_DNA
English
0
6
6
196
AgentDNA retweetledi
KC
KC@KCRubix·
Insanely accurate @RaoulGMI. Economic singularity is almost here. 1. Would prefer Universal Basic Entreprenership can also be added to UBE (owning the substrate + creating the substrate). Everyone can have a shot at it. 2. Proof of human becomes highly valuable. L1s become the only arbiter of UBE. 3. Building p2p graph @RubixChain for 1 bn tps agentic commerce/tokenization and @Agent_DNA for trust resolution.
Raoul Pal@RaoulGMI

Forget UBI. The answer is Universal Basic Equity… and it’s humanity’s pension plan for the post-AGI world... The Economic Singularity is coming faster than people think and the default question is how humans make money in a world that doesn’t really need them anymore. The default answer is UBI, which is transfer payments from a state, funded by taxing an AI economy that nation states can neither see nor keep up with. It’s a 20th century answer to a 21st century problem and it’s broken before it even starts. Agents are becoming the dominant user of the internet, not humans. Your AI is becoming your entire front end UX. The clicks economy is dying everywhere except where humans pay to feel something - clothing, travel, luxury, experiences, culture. Agents run on crypto rails because nothing else works. The dollar doesn’t fractionalise below a cent, settlement isn’t instant, permissions are required, jurisdictions matter. Stablecoins handle the dollar leg and native tokens handle the rest. The biggest users of DeFi in five years won’t be humans farming yield… it’ll be agents managing treasuries, swapping, earning and spending at machine speed. Capital formation has already shown its new shape and it came from the most unexpected place. Memecoins. Everyone wrote them off as a casino but they were a prototype. Instant capital formation around the attention of an idea, raised by entities without legal personhood, settled in seconds. That is the template agent economies will use to fund themselves. And it’s not just agents... Robots will run on the same rails, with zk permissions issued from our wallets as the source of truth, because biometrics are far too flawed for that role Open source code itself gets tokenized and finally captures the value it creates, instead of being monetized through bolted-on services and subscriptions. Proof of humanhood becomes the trust layer that lets us release agents into the world without society collapsing under synthetic noise. Identity, authentication, verification, permissioning, all of it migrates onto the same substrate. So when you zoom out, the L1s aren’t just settling agent transactions but settling the entire coordination layer of the new economy… agents, robots, humans, code, capital, identity and trust. Every contract, every treasury, every permission, every stake. Open source finally captures the value it creates, at scale, for the first time, and truly vast value accrues to the coordination layer because everything routes through it. Which brings us to the actual answer to the Economic Singularity… Universal Basic Equity. Anyone on earth with a phone and an internet connection can buy a stake in the substrate that the new economy runs on. No KYC walls, no accreditation rules, no jurisdiction, no employer, no state, no permission. The first homogenous, permissionless, globally fractionalisable claim on the productive infrastructure of the world. It's not a slogan but a structural fact about how blockchains actually work. This is their purpose. Wealth comes from owning the substrate. Income comes from being human, because attention and experience remain the irreducible currency of culture, community and love. Abundance of goods and services from AI handles the cost of living. Taxing data center electricity use solves the tax issue. Four legs of a stool that holds up the post-singularity human world. So… just buy the fucking tokens. Bitcoin if you want pure store of value, a basket of the major L1s if you want the coordination layer. 10% of your earnings, every month, for a decade. You'll be wealthy and protected from the changes to come. Crypto is going to $100trn in the next 6 to 8 years and well beyond that after. You can choose to invest in your own economic disruption, or get left behind by it. And if you’re worried about timing the cycle… …adjust your time horizon. This is humanity’s pension plan. It's all so absurdly fucking obvious...

English
0
5
5
90
AgentDNA retweetledi
KC
KC@KCRubix·
Generously scoped API permissions - no runtime authorization. No partitioned backups. CLI tokens can easily be accessed across. Your internal agent can be a big threat. Avoid these pitfalls by building agent lineage, Chain of Custody and immutability with @Agent_DNA for better security. @Railway @cursor_ai
JER@lifeof_jer

x.com/i/article/2048…

English
0
4
5
107
AgentDNA retweetledi
_Graph
_Graph@RubixChain·
Existing enterprise platforms charging per seat offering different vertical solutions would not deliver productivity to enterprises deploying Agents. Agents do need an implementation and trust layer, agree, but that would more look like a full automated trust layer like @RubixChain and @Agent_DNA
English
0
1
1
19
AgentDNA retweetledi
KC
KC@KCRubix·
Federated OAuth access (like Google OAuth) carries significsnt risks. This will get worse with AI Agents. Switch to more secure access alternatives like @ThinC_AUTH and @Agent_DNA
Peer Richelsen@peer_rich

OAuth scams are becoming more and more popular this time a malicious @nylas OAuth client is being used to attack one of our employees (who immediately understood its a scam and has reported it) the playbook is simple: 1. obtain a Google OAuth client 2. pretend to send a .docx 3. immediately open the Google OAuth splash screen 4. Attacker gains full access to Email, Calendar and more officially it falls under phishing, however attackers no longer make fake login forms but instead show official google log in screens most people fall for it because they check the website, see google.⁠com AND nylas.⁠com (which is also legit) and think its safe. signing into google to view a document is also very common

English
0
3
4
103
AgentDNA retweetledi
KC
KC@KCRubix·
@BenBajarin Historically, whenever memory capex ramped hard, cycle peaked. AI super cycle though is keeping the cycle longer. Memory scaling and contrxt optimization through SW tools like @RubixChain and @TrieNetwork will play a key role in easing memory demand growth this time.
English
0
3
4
286
AgentDNA retweetledi
_Graph
_Graph@RubixChain·
The #AGIBarbell. Open, shared and decentralized software will play a big role on the AI last mile, for security, context and memory. @TrieNetwork and @Agent_DNA layers on Rubix will balance the #barbell.
_Graph tweet media
English
1
11
4
166
AgentDNA retweetledi
Daniel San
Daniel San@dani_avila7·
Been testing Claude Managed Agents Here's my feedback: Vaults store OAuth tokens outside the sandbox, the agent never handles them directly The problem, vaults are workspace-scoped Anyone with workspace access, via API key or the Console, can reference your vaults and use your credentials in their own sessions
Daniel San tweet media
English
6
5
36
4.9K
AgentDNA retweetledi
KC
KC@KCRubix·
@chiochioball @dani_avila7 Trust boundaries, not capability, is the key adoption blocker, absolutely. Immutable, verifiable audit logs too, because agents are known to alter the logs. @Agent_DNA
English
1
2
3
38
AgentDNA retweetledi
KC
KC@KCRubix·
@schneierblog Quite informative @schneierblog. The attackers will definitely move up the stack, which has been the trend even before Mythos. Identity determination will be more difficult. The intelligent agent can be poisoned or infected by attackers using the same AI model.
English
1
3
4
415
AgentDNA retweetledi
KC
KC@KCRubix·
@natashamalpani The Scoring Function is the right phrase. Measurable. Auditable. Automated context towards self improvement.
English
1
2
3
158