fatty retweeted
imagine you deposit money into your bank account. you check your balance immediately, but it still shows the old amount. you refresh the page, and the new balance appears. when you check from another device, you might see a different amount.
this can happen when information takes time to update across different parts of a system.
@RialoHQ consistency tutorial explains six ways a system can handle this and what users should expect when they read data.
๐ญ. ๐๐๐ฟ๐ผ๐ป๐ด ๐ฐ๐ผ๐ป๐๐ถ๐๐๐ฒ๐ป๐ฐ๐
every read returns the most recently written value. this helps ensure that users see the latest data when they request it.
๐ฎ. ๐ฏ๐ผ๐๐ป๐ฑ๐ฒ๐ฑ ๐๐๐ฎ๐น๐ฒ๐ป๐ฒ๐๐
a read may return older data, but only within a set limit. this limit can be based on time or the number of updates, helping keep delays predictable.
๐ฏ. ๐ฐ๐ผ๐ป๐๐ถ๐๐๐ฒ๐ป๐ ๐ฝ๐ฟ๐ฒ๐ณ๐ถ๐
updates appear in the correct order, even when a user has not seen the latest ones. for example, if three updates happen in sequence, a user will not see the third update without seeing the first two.
๐ฐ. ๐บ๐ผ๐ป๐ผ๐๐ผ๐ป๐ถ๐ฐ ๐ฟ๐ฒ๐ฎ๐ฑ๐
Once you have seen a particular value, your later reads will not show an older one. your view of the data will not move backwards, even if it is not fully up to date.
๐ฑ. ๐ฟ๐ฒ๐ฎ๐ฑ ๐บ๐ ๐๐ฟ๐ถ๐๐ฒ๐
after you make an update, your subsequent reads will show that update or a newer one. for example, if you change your profile name, you should see the new name when you check again.
๐ฒ. ๐ฒ๐๐ฒ๐ป๐๐๐ฎ๐น ๐ฐ๐ผ๐ป๐๐ถ๐๐๐ฒ๐ป๐ฐ๐
different copies of the data may temporarily show different values. if updates stop, the copies will eventually converge and show the same value.
๐ต๐ผ๐ ๐ฑ๐ผ ๐๐ต๐ฒ๐๐ฒ ๐ด๐๐ฎ๐ฟ๐ฎ๐ป๐๐ฒ๐ฒ๐ ๐ฑ๐ถ๐ณ๐ณ๐ฒ๐ฟ?
each guarantee offers a different way of handling updates and reads.
strong consistency provides the strongest guarantee, while eventual consistency provides the weakest. the four guarantees in between offer different trade-offs and are not strictly ordered by strength.
for example, a system handling financial transactions may need to return the latest information. a system displaying view counts may be able to tolerate a short delay before every copy reflects an update.
the choice depends on how an application uses its data and what level of freshness it needs.
rialo consistency tutorial helps explain these differences and how consistency, availability and response time affect the way distributed systems handle information.
grialo
@ericargent31113
@dfwdora
@0x_alextine
imagine you deposit money into your bank account. you check your balance immediately, but it still shows the old amount. you refresh the page, and the new balance appears. when you check from another device, you might see a different amount.
this can happen when information takes time to update across different parts of a system.
@RialoHQ consistency tutorial explains six ways a system can handle this and what users should expect when they read data.
๐ญ. ๐๐๐ฟ๐ผ๐ป๐ด ๐ฐ๐ผ๐ป๐๐ถ๐๐๐ฒ๐ป๐ฐ๐
every read returns the most recently written value. this helps ensure that users see the latest data when they request it.
๐ฎ. ๐ฏ๐ผ๐๐ป๐ฑ๐ฒ๐ฑ ๐๐๐ฎ๐น๐ฒ๐ป๐ฒ๐๐
a read may return older data, but only within a set limit. this limit can be based on time or the number of updates, helping keep delays predictable.
๐ฏ. ๐ฐ๐ผ๐ป๐๐ถ๐๐๐ฒ๐ป๐ ๐ฝ๐ฟ๐ฒ๐ณ๐ถ๐
updates appear in the correct order, even when a user has not seen the latest ones. for example, if three updates happen in sequence, a user will not see the third update without seeing the first two.
๐ฐ. ๐บ๐ผ๐ป๐ผ๐๐ผ๐ป๐ถ๐ฐ ๐ฟ๐ฒ๐ฎ๐ฑ๐
Once you have seen a particular value, your later reads will not show an older one. your view of the data will not move backwards, even if it is not fully up to date.
๐ฑ. ๐ฟ๐ฒ๐ฎ๐ฑ ๐บ๐ ๐๐ฟ๐ถ๐๐ฒ๐
after you make an update, your subsequent reads will show that update or a newer one. for example, if you change your profile name, you should see the new name when you check again.
๐ฒ. ๐ฒ๐๐ฒ๐ป๐๐๐ฎ๐น ๐ฐ๐ผ๐ป๐๐ถ๐๐๐ฒ๐ป๐ฐ๐
different copies of the data may temporarily show different values. if updates stop, the copies will eventually converge and show the same value.
๐ต๐ผ๐ ๐ฑ๐ผ ๐๐ต๐ฒ๐๐ฒ ๐ด๐๐ฎ๐ฟ๐ฎ๐ป๐๐ฒ๐ฒ๐ ๐ฑ๐ถ๐ณ๐ณ๐ฒ๐ฟ?
each guarantee offers a different way of handling updates and reads.
strong consistency provides the strongest guarantee, while eventual consistency provides the weakest. the four guarantees in between offer different trade-offs and are not strictly ordered by strength.
for example, a system handling financial transactions may need to return the latest information. a system displaying view counts may be able to tolerate a short delay before every copy reflects an update.
the choice depends on how an application uses its data and what level of freshness it needs.
rialo consistency tutorial helps explain these differences and how consistency, availability and response time affect the way distributed systems handle information.
grialo
@ericargent31113
@dfwdora
@0x_alextine
fatty retweeted
A blockchain can have thousands of validators and still depend on an outside service to decide when an action should begin.
๐๐ฐ๐ธ ๐ช๐ด ๐ต๐ฉ๐ข๐ต ๐ฑ๐ฐ๐ด๐ด๐ช๐ฃ๐ญ๐ฆ?
To understand this, we need to look at how a blockchain agrees on transactions and how those transactions get started.
๐ช๐ต๐ฎ๐ ๐ฑ๐ผ๐ฒ๐ ๐ฐ๐ผ๐ป๐๐ฒ๐ป๐๐๐ ๐ฑ๐ผ?
Consensus allows validators to agree on transactions and the order in which they are recorded on the blockchain.
Having many validators distributes this responsibility, but consensus does not automatically initiate actions when conditions are met. something still needs to detect those conditions and submit the transaction.
๐๐ผ๐ป๐๐ถ๐ฑ๐ฒ๐ฟ ๐ฎ ๐น๐ฒ๐ป๐ฑ๐ถ๐ป๐ด ๐ฝ๐ฟ๐ผ๐๐ผ๐ฐ๐ผ๐น.
Imagine someone borrows money and uses their cryptocurrency as a promise to pay it back. if the value of that crypto drops too much, some of it may be sold to help repay the loan. this is called liquidation.
The lending protocol can check the value of the crypto to see if it has dropped too low. a keeper bot, which is a program that watches for these changes and sends transactions, may then start the liquidation process.
Until the transaction is sent and processed, the crypto may be at risk of being sold, but the liquidation has not yet happened.
@fermah_xyz kernel allows protocols to run sequences of actions when specific conditions are met.
Three concepts help explain how this relates to protocol agency:
โ๐๐ผ๐ป๐๐ฒ๐ป๐๐๐: validators agree on transactions and the order in which they are recorded.
โ ๐๐ฐ๐๐ถ๐ผ๐ป ๐ถ๐ป๐ถ๐๐ถ๐ฎ๐๐ถ๐ผ๐ป: a transaction needs to be submitted for the next action to begin. This may depend on an external service.
โ ๐ฃ๐ฟ๐ผ๐๐ผ๐ฐ๐ผ๐น ๐ฎ๐ด๐ฒ๐ป๐ฐ๐: a protocol can initiate actions when specific conditions are met, allowing a process to move from one step to the next according to its rules.
For example, when a lending position reaches its liquidation threshold, that condition can trigger the next step in the process. kernel also supports connecting actions across protocols, allowing an action in one protocol to lead to further actions in another.
๐ช๐ต๐ ๐ฑ๐ผ๐ฒ๐ ๐๐ต๐ถ๐ ๐บ๐ฎ๐๐๐ฒ๐ฟ ๐ณ๐ผ๐ฟ ๐ฑ๐ฒ๐ฐ๐ฒ๐ป๐๐ฟ๐ฎ๐น๐ถ๐๐ฎ๐๐ถ๐ผ๐ป?
If a protocol relies on an external service to initiate important operations, delays or failures in that service can affect when those operations happen.
A blockchain may be decentralized in how it agrees on transactions while still depending on an outside operator to initiate certain actions.
Kernel connects conditions to actions, reducing reliance on external services to start every step and giving protocols more control over how actions begin and continue.
Consensus determines which transactions are agreed on and their order, while protocol agency focuses on starting actions when the required conditions are met. both help blockchains respond to changes and carry out actions automatically.
A decentralized blockchain should not only agree on what happens. It should also have a reliable way to initiate the next action when the required conditions are met.
@7wealthh
A chain can have thousands of validators and still have a centralized calendar.
Fermah enables onchain conditions to start and continue actions across protocols.
Consensus decides order. Protocol agency decides when action begins.
Decentralized ordering without decentralized initiation leaves protocol agency outside the protocol.
A blockchain can have thousands of validators and still depend on an outside service to decide when an action should begin.
๐๐ฐ๐ธ ๐ช๐ด ๐ต๐ฉ๐ข๐ต ๐ฑ๐ฐ๐ด๐ด๐ช๐ฃ๐ญ๐ฆ?
To understand this, we need to look at how a blockchain agrees on transactions and how those transactions get started.
๐ช๐ต๐ฎ๐ ๐ฑ๐ผ๐ฒ๐ ๐ฐ๐ผ๐ป๐๐ฒ๐ป๐๐๐ ๐ฑ๐ผ?
Consensus allows validators to agree on transactions and the order in which they are recorded on the blockchain.
Having many validators distributes this responsibility, but consensus does not automatically initiate actions when conditions are met. something still needs to detect those conditions and submit the transaction.
๐๐ผ๐ป๐๐ถ๐ฑ๐ฒ๐ฟ ๐ฎ ๐น๐ฒ๐ป๐ฑ๐ถ๐ป๐ด ๐ฝ๐ฟ๐ผ๐๐ผ๐ฐ๐ผ๐น.
Imagine someone borrows money and uses their cryptocurrency as a promise to pay it back. if the value of that crypto drops too much, some of it may be sold to help repay the loan. this is called liquidation.
The lending protocol can check the value of the crypto to see if it has dropped too low. a keeper bot, which is a program that watches for these changes and sends transactions, may then start the liquidation process.
Until the transaction is sent and processed, the crypto may be at risk of being sold, but the liquidation has not yet happened.
@fermah_xyz kernel allows protocols to run sequences of actions when specific conditions are met.
Three concepts help explain how this relates to protocol agency:
โ๐๐ผ๐ป๐๐ฒ๐ป๐๐๐: validators agree on transactions and the order in which they are recorded.
โ ๐๐ฐ๐๐ถ๐ผ๐ป ๐ถ๐ป๐ถ๐๐ถ๐ฎ๐๐ถ๐ผ๐ป: a transaction needs to be submitted for the next action to begin. This may depend on an external service.
โ ๐ฃ๐ฟ๐ผ๐๐ผ๐ฐ๐ผ๐น ๐ฎ๐ด๐ฒ๐ป๐ฐ๐: a protocol can initiate actions when specific conditions are met, allowing a process to move from one step to the next according to its rules.
For example, when a lending position reaches its liquidation threshold, that condition can trigger the next step in the process. kernel also supports connecting actions across protocols, allowing an action in one protocol to lead to further actions in another.
๐ช๐ต๐ ๐ฑ๐ผ๐ฒ๐ ๐๐ต๐ถ๐ ๐บ๐ฎ๐๐๐ฒ๐ฟ ๐ณ๐ผ๐ฟ ๐ฑ๐ฒ๐ฐ๐ฒ๐ป๐๐ฟ๐ฎ๐น๐ถ๐๐ฎ๐๐ถ๐ผ๐ป?
If a protocol relies on an external service to initiate important operations, delays or failures in that service can affect when those operations happen.
A blockchain may be decentralized in how it agrees on transactions while still depending on an outside operator to initiate certain actions.
Kernel connects conditions to actions, reducing reliance on external services to start every step and giving protocols more control over how actions begin and continue.
Consensus determines which transactions are agreed on and their order, while protocol agency focuses on starting actions when the required conditions are met. both help blockchains respond to changes and carry out actions automatically.
A decentralized blockchain should not only agree on what happens. It should also have a reliable way to initiate the next action when the required conditions are met.
@7wealthh
A chain can have thousands of validators and still have a centralized calendar.
Fermah enables onchain conditions to start and continue actions across protocols.
Consensus decides order. Protocol agency decides when action begins.
Decentralized ordering without decentralized initiation leaves protocol agency outside the protocol.
fatty retweeted
websites contain different kinds of information, from account balances and trading activity to identity details and transaction records.
when an application needs information from an online service, it must identify the source and specify the data it wants to retrieve and verify.
for example, a financial application might need trading information from a cryptocurrency exchange, while another application might need different information from another online service.
๐ฉ๐ฐ๐ธ ๐ค๐ข๐ฏ ๐ต๐ฉ๐ฆ๐ด๐ฆ ๐ฅ๐ข๐ต๐ข ๐ณ๐ฆ๐ฒ๐ถ๐ช๐ณ๐ฆ๐ฎ๐ฆ๐ฏ๐ต๐ด ๐ฃ๐ฆ ๐ฐ๐ณ๐จ๐ข๐ฏ๐ช๐ด๐ฆ๐ฅ?
@primus_labs provides data templates to help organise information from sources outside the blockchain.
๐ช๐ต๐ฎ๐ ๐ฎ๐ฟ๐ฒ ๐ฑ๐ฎ๐๐ฎ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ๐?
data templates are predefined formats used to represent information from off-chain sources. they help organise how applications request specific information for verification.
through the ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐ฑ๐ฒ๐๐ฒ๐น๐ผ๐ฝ๐ฒ๐ฟ ๐ต๐๐ฏ, users can find existing templates or create their own to suit their requirements.
๐๐ผ๐ ๐ฑ๐ผ ๐๐ต๐ฒ๐ ๐๐ผ๐ฟ๐ธ?
the process involves three main steps.
๐ญ. ๐ถ๐ฑ๐ฒ๐ป๐๐ถ๐ณ๐ ๐๐ต๐ฒ ๐ฑ๐ฎ๐๐ฎ ๐๐ผ๐๐ฟ๐ฐ๐ฒ
first, identify the online service that contains the information the application needs.
๐ฎ. ๐๐ฒ๐น๐ฒ๐ฐ๐ ๐ผ๐ฟ ๐ฐ๐ฟ๐ฒ๐ฎ๐๐ฒ ๐ฎ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ
choose an existing template from the ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐ฑ๐ฒ๐๐ฒ๐น๐ผ๐ฝ๐ฒ๐ฟ ๐ต๐๐ฏ or create one that suits the required data source and information.
๐ฏ. ๐๐๐ฒ ๐๐ต๐ฒ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ ๐๐
each template has a unique ID. an application can use this ID when interacting with primus network attestors to carry out data verification tasks.
๐ช๐ต๐ ๐ฎ๐ฟ๐ฒ ๐ฑ๐ฎ๐๐ฎ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ๐ ๐๐๐ฒ๐ณ๐๐น?
different online services provide different kinds of information. data templates give users a structured way to define data sources and prepare them for use in applications.
primus data templates work alongside its ๐๐ผ๐ณ๐๐๐ฎ๐ฟ๐ฒ ๐ฑ๐ฒ๐๐ฒ๐น๐ผ๐ฝ๐บ๐ฒ๐ป๐ ๐ธ๐ถ๐๐ (SDKs), which provide tools needed to integrate the verification process into applications.
together, these components help applications verify information from online services.
fatty retweeted
Been looking into @conso_xyz lately, and I think the idea behind it is interesting.
In this video, I break down what CONSO is building, the consumer passport, and how it wants to make our digital activities and reputation more valuable across different platforms.
What do you think about this?
websites contain different kinds of information, from account balances and trading activity to identity details and transaction records.
when an application needs information from an online service, it must identify the source and specify the data it wants to retrieve and verify.
for example, a financial application might need trading information from a cryptocurrency exchange, while another application might need different information from another online service.
๐ฉ๐ฐ๐ธ ๐ค๐ข๐ฏ ๐ต๐ฉ๐ฆ๐ด๐ฆ ๐ฅ๐ข๐ต๐ข ๐ณ๐ฆ๐ฒ๐ถ๐ช๐ณ๐ฆ๐ฎ๐ฆ๐ฏ๐ต๐ด ๐ฃ๐ฆ ๐ฐ๐ณ๐จ๐ข๐ฏ๐ช๐ด๐ฆ๐ฅ?
@primus_labs provides data templates to help organise information from sources outside the blockchain.
๐ช๐ต๐ฎ๐ ๐ฎ๐ฟ๐ฒ ๐ฑ๐ฎ๐๐ฎ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ๐?
data templates are predefined formats used to represent information from off-chain sources. they help organise how applications request specific information for verification.
through the ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐ฑ๐ฒ๐๐ฒ๐น๐ผ๐ฝ๐ฒ๐ฟ ๐ต๐๐ฏ, users can find existing templates or create their own to suit their requirements.
๐๐ผ๐ ๐ฑ๐ผ ๐๐ต๐ฒ๐ ๐๐ผ๐ฟ๐ธ?
the process involves three main steps.
๐ญ. ๐ถ๐ฑ๐ฒ๐ป๐๐ถ๐ณ๐ ๐๐ต๐ฒ ๐ฑ๐ฎ๐๐ฎ ๐๐ผ๐๐ฟ๐ฐ๐ฒ
first, identify the online service that contains the information the application needs.
๐ฎ. ๐๐ฒ๐น๐ฒ๐ฐ๐ ๐ผ๐ฟ ๐ฐ๐ฟ๐ฒ๐ฎ๐๐ฒ ๐ฎ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ
choose an existing template from the ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐ฑ๐ฒ๐๐ฒ๐น๐ผ๐ฝ๐ฒ๐ฟ ๐ต๐๐ฏ or create one that suits the required data source and information.
๐ฏ. ๐๐๐ฒ ๐๐ต๐ฒ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ ๐๐
each template has a unique ID. an application can use this ID when interacting with primus network attestors to carry out data verification tasks.
๐ช๐ต๐ ๐ฎ๐ฟ๐ฒ ๐ฑ๐ฎ๐๐ฎ ๐๐ฒ๐บ๐ฝ๐น๐ฎ๐๐ฒ๐ ๐๐๐ฒ๐ณ๐๐น?
different online services provide different kinds of information. data templates give users a structured way to define data sources and prepare them for use in applications.
primus data templates work alongside its ๐๐ผ๐ณ๐๐๐ฎ๐ฟ๐ฒ ๐ฑ๐ฒ๐๐ฒ๐น๐ผ๐ฝ๐บ๐ฒ๐ป๐ ๐ธ๐ถ๐๐ (SDKs), which provide tools needed to integrate the verification process into applications.
together, these components help applications verify information from online services.
fatty retweeted
๐ฃ๐ฟ๐ถ๐บ๐๐ ๐ฒ๐
๐ฝ๐น๐ฎ๐ถ๐ป๐ฒ๐ฑ: ๐๐ฒ๐ฟ๐ถ๐ณ๐๐ถ๐ป๐ด ๐ถ๐ป๐๐ฒ๐ฟ๐ป๐ฒ๐ ๐ฑ๐ฎ๐๐ฎ ๐ฎ๐ป๐ฑ ๐ฐ๐ผ๐บ๐ฝ๐๐๐ถ๐ป๐ด ๐ฝ๐ฟ๐ถ๐๐ฎ๐๐ฒ๐น๐
every day, we use websites and online services that hold information about us. this can include financial records, account activity, identity details, and trading history.
sometimes, an application needs this information to make a decision, but sharing all your personal details just to prove one thing can put your privacy at risk.
๐ธ๐ฉ๐ข๐ต ๐ช๐ง ๐บ๐ฐ๐ถ ๐ค๐ฐ๐ถ๐ญ๐ฅ ๐ฑ๐ณ๐ฐ๐ท๐ฆ ๐ต๐ฉ๐ข๐ต ๐ด๐ฐ๐ฎ๐ฆ๐ต๐ฉ๐ช๐ฏ๐จ ๐ช๐ด ๐ต๐ณ๐ถ๐ฆ ๐ธ๐ช๐ต๐ฉ๐ฐ๐ถ๐ต ๐ณ๐ฆ๐ท๐ฆ๐ข๐ญ๐ช๐ฏ๐จ ๐ข๐ญ๐ญ ๐ต๐ฉ๐ฆ ๐ฅ๐ฆ๐ต๐ข๐ช๐ญ๐ด ๐ฃ๐ฆ๐ฉ๐ช๐ฏ๐ฅ ๐ช๐ต?
@primus_labs is working on technologies that help applications verify information from the internet while protecting sensitive data.
primus is a trust, verification, and privacy layer for the crypto and AI agent economy.
๐๐ผ๐ ๐ฑ๐ผ๐ฒ๐ ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐๐ผ๐ฟ๐ธ?
primus uses two key technologies:
๐ญ. ๐๐ธ๐ง๐๐ฆ (๐๐ฒ๐ฟ๐ผ-๐ธ๐ป๐ผ๐๐น๐ฒ๐ฑ๐ด๐ฒ ๐๐ฟ๐ฎ๐ป๐๐ฝ๐ผ๐ฟ๐ ๐น๐ฎ๐๐ฒ๐ฟ ๐๐ฒ๐ฐ๐๐ฟ๐ถ๐๐): verifying information from websites
when you visit a website, TLS helps protect the information exchanged between your browser and that website.
primus builds on TLS by using cryptographic proofs to verify the authenticity of information from online services without revealing all the underlying data.
for example, imagine a crypto lending platform needs to check information about your assets. you could provide proof of relevant information without making your entire financial history public.
๐ฎ. ๐๐ธ๐๐๐ (๐๐ฒ๐ฟ๐ผ-๐ธ๐ป๐ผ๐๐น๐ฒ๐ฑ๐ด๐ฒ ๐ณ๐๐น๐น๐ ๐ต๐ผ๐บ๐ผ๐บ๐ผ๐ฟ๐ฝ๐ต๐ถ๐ฐ ๐ฒ๐ป๐ฐ๐ฟ๐๐ฝ๐๐ถ๐ผ๐ป): working with encrypted data
sometimes, an application needs to do more than check information. it needs to calculate something using that information.
๐ฉ๐ฐ๐ธ ๐ค๐ข๐ฏ ๐ค๐ข๐ญ๐ค๐ถ๐ญ๐ข๐ต๐ช๐ฐ๐ฏ๐ด ๐ฃ๐ฆ ๐ฑ๐ฆ๐ณ๐ง๐ฐ๐ณ๐ฎ๐ฆ๐ฅ ๐ธ๐ช๐ต๐ฉ๐ฐ๐ถ๐ต ๐ฆ๐น๐ฑ๐ฐ๐ด๐ช๐ฏ๐จ ๐ต๐ฉ๐ฆ ๐ฐ๐ณ๐ช๐จ๐ช๐ฏ๐ข๐ญ ๐ฅ๐ข๐ต๐ข?
fully homomorphic encryption (FHE) allows calculations on encrypted data without decrypting it first. the result remains encrypted until an authorised party decrypts it.
primus uses zkFHE to combine FHE with zero-knowledge techniques, helping verify that calculations on encrypted data were performed correctly.
๐ธ๐ฉ๐ฆ๐ณ๐ฆ ๐ค๐ข๐ฏ ๐ต๐ฉ๐ฆ๐ด๐ฆ ๐ต๐ฆ๐ค๐ฉ๐ฏ๐ฐ๐ญ๐ฐ๐จ๐ช๐ฆ๐ด ๐ฃ๐ฆ ๐ถ๐ด๐ฆ๐ง๐ถ๐ญ?
here are a few examples:
โ ๐ฐ๐ฟ๐๐ฝ๐๐ผ ๐น๐ฒ๐ป๐ฑ๐ถ๐ป๐ด: a lending platform could check a userโs assets without asking them to share their full financial history.
โ ๐ถ๐ฑ๐ฒ๐ป๐๐ถ๐๐ ๐๐ฒ๐ฟ๐ถ๐ณ๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป: an application could verify specific information about a person without requiring them to reveal unnecessary personal details.
โ ๐๐ ๐ฎ๐ด๐ฒ๐ป๐๐: primus can help verify that an AI agent made the required requests to a website and received the expected responses while protecting private information.
these are some ways primus technologies can be applied, depending on how each application uses them.
๐ช๐ต๐ ๐ฑ๐ผ๐ฒ๐ ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐บ๐ฎ๐๐๐ฒ๐ฟ?
much of the information used by blockchain applications comes from outside the blockchain. they need ways to verify it without exposing private details.
primus brings together two important capabilities:
โ ๐ฑ๐ฎ๐๐ฎ ๐๐ฒ๐ฟ๐ถ๐ณ๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป: verifying the authenticity of internet data without revealing the data itself.
โ ๐ฝ๐ฟ๐ถ๐๐ฎ๐ฐ๐-๐ฝ๐ฟ๐ฒ๐๐ฒ๐ฟ๐๐ถ๐ป๐ด ๐ฐ๐ผ๐บ๐ฝ๐๐๐ฎ๐๐ถ๐ผ๐ป: allowing calculations on encrypted data while helping verify that the results are correct.
primus aims to make these capabilities available to developers building applications across blockchain, finance, and AI.
fatty retweeted
some people are about to walk away with juicy rewards from $AXIS.
Iโm talking about those who positioned themselves across multiple angles.
I know a few people who are well positioned in:
- Axis Hub
- Binance campaign
- Discord roles
- Kaito campaign
if $AXIS launches at a decent FDV, some of these people could be looking at a solid 4-figure or even 5-figure airdrop.
how many angles are you positioned on?
๐ค Made with AI
$AXIS might be closer than we expect, and hereโs why I think so.
- discord roles capped and canโt be earned anymore.
- 2 days ago, the Axis community sale wrapped up. the sale allocation was revealed and refunds were sent to participants swiftly.
- 52 days ago, the $AXIS token was minted on Base.
ca: 0x5bE75e3b0aFf5D1A042A5363Ae36AcB1d6d755ea
- the Binance campaign has ended.
- Kaito Epoch 2 ends in less than 3 weeks, while weโre still waiting for the Epoch 1 leaderboard to be released.
putting all of these together, Iโm think the TGE might happen in the last few days of October.
nothing confirmed yet, just connecting the dots.
๐ฃ๐ฟ๐ถ๐บ๐๐ ๐ฒ๐
๐ฝ๐น๐ฎ๐ถ๐ป๐ฒ๐ฑ: ๐๐ฒ๐ฟ๐ถ๐ณ๐๐ถ๐ป๐ด ๐ถ๐ป๐๐ฒ๐ฟ๐ป๐ฒ๐ ๐ฑ๐ฎ๐๐ฎ ๐ฎ๐ป๐ฑ ๐ฐ๐ผ๐บ๐ฝ๐๐๐ถ๐ป๐ด ๐ฝ๐ฟ๐ถ๐๐ฎ๐๐ฒ๐น๐
every day, we use websites and online services that hold information about us. this can include financial records, account activity, identity details, and trading history.
sometimes, an application needs this information to make a decision, but sharing all your personal details just to prove one thing can put your privacy at risk.
๐ธ๐ฉ๐ข๐ต ๐ช๐ง ๐บ๐ฐ๐ถ ๐ค๐ฐ๐ถ๐ญ๐ฅ ๐ฑ๐ณ๐ฐ๐ท๐ฆ ๐ต๐ฉ๐ข๐ต ๐ด๐ฐ๐ฎ๐ฆ๐ต๐ฉ๐ช๐ฏ๐จ ๐ช๐ด ๐ต๐ณ๐ถ๐ฆ ๐ธ๐ช๐ต๐ฉ๐ฐ๐ถ๐ต ๐ณ๐ฆ๐ท๐ฆ๐ข๐ญ๐ช๐ฏ๐จ ๐ข๐ญ๐ญ ๐ต๐ฉ๐ฆ ๐ฅ๐ฆ๐ต๐ข๐ช๐ญ๐ด ๐ฃ๐ฆ๐ฉ๐ช๐ฏ๐ฅ ๐ช๐ต?
@primus_labs is working on technologies that help applications verify information from the internet while protecting sensitive data.
primus is a trust, verification, and privacy layer for the crypto and AI agent economy.
๐๐ผ๐ ๐ฑ๐ผ๐ฒ๐ ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐๐ผ๐ฟ๐ธ?
primus uses two key technologies:
๐ญ. ๐๐ธ๐ง๐๐ฆ (๐๐ฒ๐ฟ๐ผ-๐ธ๐ป๐ผ๐๐น๐ฒ๐ฑ๐ด๐ฒ ๐๐ฟ๐ฎ๐ป๐๐ฝ๐ผ๐ฟ๐ ๐น๐ฎ๐๐ฒ๐ฟ ๐๐ฒ๐ฐ๐๐ฟ๐ถ๐๐): verifying information from websites
when you visit a website, TLS helps protect the information exchanged between your browser and that website.
primus builds on TLS by using cryptographic proofs to verify the authenticity of information from online services without revealing all the underlying data.
for example, imagine a crypto lending platform needs to check information about your assets. you could provide proof of relevant information without making your entire financial history public.
๐ฎ. ๐๐ธ๐๐๐ (๐๐ฒ๐ฟ๐ผ-๐ธ๐ป๐ผ๐๐น๐ฒ๐ฑ๐ด๐ฒ ๐ณ๐๐น๐น๐ ๐ต๐ผ๐บ๐ผ๐บ๐ผ๐ฟ๐ฝ๐ต๐ถ๐ฐ ๐ฒ๐ป๐ฐ๐ฟ๐๐ฝ๐๐ถ๐ผ๐ป): working with encrypted data
sometimes, an application needs to do more than check information. it needs to calculate something using that information.
๐ฉ๐ฐ๐ธ ๐ค๐ข๐ฏ ๐ค๐ข๐ญ๐ค๐ถ๐ญ๐ข๐ต๐ช๐ฐ๐ฏ๐ด ๐ฃ๐ฆ ๐ฑ๐ฆ๐ณ๐ง๐ฐ๐ณ๐ฎ๐ฆ๐ฅ ๐ธ๐ช๐ต๐ฉ๐ฐ๐ถ๐ต ๐ฆ๐น๐ฑ๐ฐ๐ด๐ช๐ฏ๐จ ๐ต๐ฉ๐ฆ ๐ฐ๐ณ๐ช๐จ๐ช๐ฏ๐ข๐ญ ๐ฅ๐ข๐ต๐ข?
fully homomorphic encryption (FHE) allows calculations on encrypted data without decrypting it first. the result remains encrypted until an authorised party decrypts it.
primus uses zkFHE to combine FHE with zero-knowledge techniques, helping verify that calculations on encrypted data were performed correctly.
๐ธ๐ฉ๐ฆ๐ณ๐ฆ ๐ค๐ข๐ฏ ๐ต๐ฉ๐ฆ๐ด๐ฆ ๐ต๐ฆ๐ค๐ฉ๐ฏ๐ฐ๐ญ๐ฐ๐จ๐ช๐ฆ๐ด ๐ฃ๐ฆ ๐ถ๐ด๐ฆ๐ง๐ถ๐ญ?
here are a few examples:
โ ๐ฐ๐ฟ๐๐ฝ๐๐ผ ๐น๐ฒ๐ป๐ฑ๐ถ๐ป๐ด: a lending platform could check a userโs assets without asking them to share their full financial history.
โ ๐ถ๐ฑ๐ฒ๐ป๐๐ถ๐๐ ๐๐ฒ๐ฟ๐ถ๐ณ๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป: an application could verify specific information about a person without requiring them to reveal unnecessary personal details.
โ ๐๐ ๐ฎ๐ด๐ฒ๐ป๐๐: primus can help verify that an AI agent made the required requests to a website and received the expected responses while protecting private information.
these are some ways primus technologies can be applied, depending on how each application uses them.
๐ช๐ต๐ ๐ฑ๐ผ๐ฒ๐ ๐ฝ๐ฟ๐ถ๐บ๐๐ ๐บ๐ฎ๐๐๐ฒ๐ฟ?
much of the information used by blockchain applications comes from outside the blockchain. they need ways to verify it without exposing private details.
primus brings together two important capabilities:
โ ๐ฑ๐ฎ๐๐ฎ ๐๐ฒ๐ฟ๐ถ๐ณ๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป: verifying the authenticity of internet data without revealing the data itself.
โ ๐ฝ๐ฟ๐ถ๐๐ฎ๐ฐ๐-๐ฝ๐ฟ๐ฒ๐๐ฒ๐ฟ๐๐ถ๐ป๐ด ๐ฐ๐ผ๐บ๐ฝ๐๐๐ฎ๐๐ถ๐ผ๐ป: allowing calculations on encrypted data while helping verify that the results are correct.
primus aims to make these capabilities available to developers building applications across blockchain, finance, and AI.
fatty retweeted
๐ฌ๐ผ๐ ๐ฎ๐๐ธ๐ฒ๐ฑ ๐ฎ๐ป ๐๐ ๐ฎ๐ด๐ฒ๐ป๐ ๐๐ผ ๐ผ๐ฟ๐ฑ๐ฒ๐ฟ ๐๐ป๐ฒ๐ฎ๐ธ๐ฒ๐ฟ๐. ๐๐ ๐ฝ๐ฎ๐ถ๐ฑ ๐๐ต๐ฒ ๐ฟ๐ถ๐ด๐ต๐ ๐๐๐ผ๐ฟ๐ฒ, ๐๐๐ฎ๐๐ฒ๐ฑ ๐๐ถ๐๐ต๐ถ๐ป ๐๐ผ๐๐ฟ ๐ฏ๐๐ฑ๐ด๐ฒ๐, ๐ฏ๐๐ ๐ฏ๐ผ๐๐ด๐ต๐ ๐ฎ ๐ฏ๐ฎ๐ด.
It followed your spending rules, but did it follow your instructions?
AI agents can help us shop, book services, and make payments. Giving them permission to use our money makes these tasks easier, ๐ฃ๐ถ๐ต ๐ฉ๐ฐ๐ธ ๐ฅ๐ฐ ๐ธ๐ฆ ๐ฌ๐ฏ๐ฐ๐ธ ๐ต๐ฉ๐ฆ๐บโ๐ณ๐ฆ ๐ข๐ค๐ต๐ถ๐ข๐ญ๐ญ๐บ ๐ฅ๐ฐ๐ช๐ฏ๐จ ๐ธ๐ฉ๐ข๐ต ๐ธ๐ฆ ๐ข๐ด๐ฌ๐ฆ๐ฅ?
Thatโs the problem @MetaMask and @GenLayer are working on through Internet Court.
๐ง๐๐ผ ๐น๐ฎ๐๐ฒ๐ฟ๐ ๐๐ผ๐ฟ๐ธ๐ถ๐ป๐ด ๐๐ผ๐ด๐ฒ๐๐ต๐ฒ๐ฟ
Think of it as two checks to help keep an AI agent on track.
โ MetaMask checks the payment.
With its Smart Accounts Kit, you can let an AI agent spend a limited amount of money without giving it your private key. You can choose which tokens it can use, who it can pay, how much it can spend, and when its permission ends.
If a payment breaks these rules, it can be stopped before the money is sent.
โ GenLayer checks whether the agent followed your instructions.
๐๐ฉ๐ข๐ต ๐ช๐ง ๐ต๐ฉ๐ฆ ๐ฑ๐ข๐บ๐ฎ๐ฆ๐ฏ๐ต ๐ฑ๐ข๐ด๐ด๐ฆ๐ด ๐ข๐ญ๐ญ ๐ต๐ฉ๐ฆ ๐ค๐ฉ๐ฆ๐ค๐ฌ๐ด ๐ฃ๐ถ๐ต ๐ต๐ฉ๐ฆ ๐ข๐จ๐ฆ๐ฏ๐ต ๐ด๐ต๐ช๐ญ๐ญ ๐ฃ๐ถ๐บ๐ด ๐ต๐ฉ๐ฆ ๐ธ๐ณ๐ฐ๐ฏ๐จ ๐ต๐ฉ๐ช๐ฏ๐จ?
GenLayerโs Internet Court can review what happened and compare the evidence with what you originally asked the agent to do. Its AI jury then decides whether the agent followed your instructions.
๐๐ฟ๐ผ๐บ ๐ฝ๐ฎ๐๐บ๐ฒ๐ป๐ ๐๐ผ ๐ฒ๐ป๐ณ๐ผ๐ฟ๐ฐ๐ฒ๐บ๐ฒ๐ป๐
The integration connects GenLayer decision-making process to MetaMask x402 payment flow, which allows software to pay for services automatically.
Hereโs how it works:
โ The AI agent requests a service and tries to pay using the permission you gave it.
โ MetaMask checks whether the payment follows your rules. If it does not, the payment can be blocked before the money moves.
โ If the payment follows those rules but may go against your instructions, GenLayer reviews evidence such as payment records, your original request, and the serviceโs response.
โ Based on the outcome, the system can allow the agent to continue, limit its permissions, revoke them, or send the case for further review.
Once the decision is final, it can be sent onchain so the relevant permission change takes effect.
This connects the decision to an action, allowing the system to change what the agent is allowed to do next.
๐ฆ๐ผ ๐ต๐ผ๐ ๐ฑ๐ผ๐ฒ๐ ๐๐ต๐ฒ ๐ท๐๐ฟ๐ ๐๐ผ๐ฟ๐ธ?
โ Five validators are selected at random, each running its own AI model.
โ One proposes a decision, while the other four review it independently and vote.
โ If there is no majority, the leader changes and the process repeats.
โ Once a majority agrees, a 30-minute appeal window opens. Anyone can appeal by staking more gas, bringing more validators into the jury.
This gives the decision another review before it becomes final.
๐ช๐ต๐ฎ๐ ๐ถ๐ณ ๐ฎ๐ป ๐๐ ๐ฎ๐ด๐ฒ๐ป๐ ๐ต๐ฎ๐ ๐ฎ๐น๐ฟ๐ฒ๐ฎ๐ฑ๐ ๐ฝ๐ฎ๐ถ๐ฑ ๐ณ๐ผ๐ฟ ๐๐ต๐ฒ ๐๐ฟ๐ผ๐ป๐ด ๐๐ต๐ถ๐ป๐ด?
Revoking the agentโs permission can stop future payments, but it cannot undo a payment that has already gone through.
GenLayer is also working on a service-payment system that holds payment until the service is checked against agreed conditions. The provider can then receive the full payment or a smaller amount, depending on whether the service met those conditions.
With MetaMask managing what an agent can spend and GenLayer checking whether it follows instructions, AI agents can act on our behalf while still being held accountable for their actions.
A case study on Internet Court by @MetaMaskDev just dropped!
@MetaMask limits what your agent can pay for, internet court by @GenLayer rules on whether it actually did what you asked.
Thank you for writing this up.
metamask.io/news/genlayer-inโฆ
๐ฌ๐ผ๐ ๐ฎ๐๐ธ๐ฒ๐ฑ ๐ฎ๐ป ๐๐ ๐ฎ๐ด๐ฒ๐ป๐ ๐๐ผ ๐ผ๐ฟ๐ฑ๐ฒ๐ฟ ๐๐ป๐ฒ๐ฎ๐ธ๐ฒ๐ฟ๐. ๐๐ ๐ฝ๐ฎ๐ถ๐ฑ ๐๐ต๐ฒ ๐ฟ๐ถ๐ด๐ต๐ ๐๐๐ผ๐ฟ๐ฒ, ๐๐๐ฎ๐๐ฒ๐ฑ ๐๐ถ๐๐ต๐ถ๐ป ๐๐ผ๐๐ฟ ๐ฏ๐๐ฑ๐ด๐ฒ๐, ๐ฏ๐๐ ๐ฏ๐ผ๐๐ด๐ต๐ ๐ฎ ๐ฏ๐ฎ๐ด.
It followed your spending rules, but did it follow your instructions?
AI agents can help us shop, book services, and make payments. Giving them permission to use our money makes these tasks easier, ๐ฃ๐ถ๐ต ๐ฉ๐ฐ๐ธ ๐ฅ๐ฐ ๐ธ๐ฆ ๐ฌ๐ฏ๐ฐ๐ธ ๐ต๐ฉ๐ฆ๐บโ๐ณ๐ฆ ๐ข๐ค๐ต๐ถ๐ข๐ญ๐ญ๐บ ๐ฅ๐ฐ๐ช๐ฏ๐จ ๐ธ๐ฉ๐ข๐ต ๐ธ๐ฆ ๐ข๐ด๐ฌ๐ฆ๐ฅ?
Thatโs the problem @MetaMask and @GenLayer are working on through Internet Court.
๐ง๐๐ผ ๐น๐ฎ๐๐ฒ๐ฟ๐ ๐๐ผ๐ฟ๐ธ๐ถ๐ป๐ด ๐๐ผ๐ด๐ฒ๐๐ต๐ฒ๐ฟ
Think of it as two checks to help keep an AI agent on track.
โ MetaMask checks the payment.
With its Smart Accounts Kit, you can let an AI agent spend a limited amount of money without giving it your private key. You can choose which tokens it can use, who it can pay, how much it can spend, and when its permission ends.
If a payment breaks these rules, it can be stopped before the money is sent.
โ GenLayer checks whether the agent followed your instructions.
๐๐ฉ๐ข๐ต ๐ช๐ง ๐ต๐ฉ๐ฆ ๐ฑ๐ข๐บ๐ฎ๐ฆ๐ฏ๐ต ๐ฑ๐ข๐ด๐ด๐ฆ๐ด ๐ข๐ญ๐ญ ๐ต๐ฉ๐ฆ ๐ค๐ฉ๐ฆ๐ค๐ฌ๐ด ๐ฃ๐ถ๐ต ๐ต๐ฉ๐ฆ ๐ข๐จ๐ฆ๐ฏ๐ต ๐ด๐ต๐ช๐ญ๐ญ ๐ฃ๐ถ๐บ๐ด ๐ต๐ฉ๐ฆ ๐ธ๐ณ๐ฐ๐ฏ๐จ ๐ต๐ฉ๐ช๐ฏ๐จ?
GenLayerโs Internet Court can review what happened and compare the evidence with what you originally asked the agent to do. Its AI jury then decides whether the agent followed your instructions.
๐๐ฟ๐ผ๐บ ๐ฝ๐ฎ๐๐บ๐ฒ๐ป๐ ๐๐ผ ๐ฒ๐ป๐ณ๐ผ๐ฟ๐ฐ๐ฒ๐บ๐ฒ๐ป๐
The integration connects GenLayer decision-making process to MetaMask x402 payment flow, which allows software to pay for services automatically.
Hereโs how it works:
โ The AI agent requests a service and tries to pay using the permission you gave it.
โ MetaMask checks whether the payment follows your rules. If it does not, the payment can be blocked before the money moves.
โ If the payment follows those rules but may go against your instructions, GenLayer reviews evidence such as payment records, your original request, and the serviceโs response.
โ Based on the outcome, the system can allow the agent to continue, limit its permissions, revoke them, or send the case for further review.
Once the decision is final, it can be sent onchain so the relevant permission change takes effect.
This connects the decision to an action, allowing the system to change what the agent is allowed to do next.
๐ฆ๐ผ ๐ต๐ผ๐ ๐ฑ๐ผ๐ฒ๐ ๐๐ต๐ฒ ๐ท๐๐ฟ๐ ๐๐ผ๐ฟ๐ธ?
โ Five validators are selected at random, each running its own AI model.
โ One proposes a decision, while the other four review it independently and vote.
โ If there is no majority, the leader changes and the process repeats.
โ Once a majority agrees, a 30-minute appeal window opens. Anyone can appeal by staking more gas, bringing more validators into the jury.
This gives the decision another review before it becomes final.
๐ช๐ต๐ฎ๐ ๐ถ๐ณ ๐ฎ๐ป ๐๐ ๐ฎ๐ด๐ฒ๐ป๐ ๐ต๐ฎ๐ ๐ฎ๐น๐ฟ๐ฒ๐ฎ๐ฑ๐ ๐ฝ๐ฎ๐ถ๐ฑ ๐ณ๐ผ๐ฟ ๐๐ต๐ฒ ๐๐ฟ๐ผ๐ป๐ด ๐๐ต๐ถ๐ป๐ด?
Revoking the agentโs permission can stop future payments, but it cannot undo a payment that has already gone through.
GenLayer is also working on a service-payment system that holds payment until the service is checked against agreed conditions. The provider can then receive the full payment or a smaller amount, depending on whether the service met those conditions.
With MetaMask managing what an agent can spend and GenLayer checking whether it follows instructions, AI agents can act on our behalf while still being held accountable for their actions.
A case study on Internet Court by @MetaMaskDev just dropped!
@MetaMask limits what your agent can pay for, internet court by @GenLayer rules on whether it actually did what you asked.
Thank you for writing this up.
metamask.io/news/genlayer-inโฆ
fatty retweeted
I turned an idea about the future of justice into charcoal art.
On one side, traditional courts represent the long waits, uncertainty, and emotional weight that can come with resolving disputes.
On the other, @courtofinternet, powered by @GenLayer, represents a new possibility: decentralized AI-powered dispute resolution designed for a world where humans and AI agents increasingly interact and do business.
The idea is simple. When two parties cannot agree, a decentralized AI jury can evaluate the evidence and help reach a verdict in minutes.
But this is bigger than speed.
As autonomous agents begin negotiating, delivering services, and exchanging value, trust becomes a critical part of the infrastructure. Smart contracts can execute rules, but disputes often require judgment.
That is the problem Internet Court is exploring.
And the exciting part? We're still early in discovering what this could unlock for builders, developers, businesses, and the wider Web3 ecosystem.
I'm interested in seeing what the community can build around this idea, from agent-to-agent commerce to new ways of handling digital agreements.
Builders, researchers, and potential ecosystem partners: what use case would you want to see explored first?
The future of the internet won't just need agents that can transact. It will need ways for them to resolve disagreements, too.
Haven't explored Internet Court yet? Take a few minutes to discover how it works, explore the idea, and join the conversation > internetcourt.org/
Your next big idea could be one of the use cases that helps shape the future of decentralized dispute resolution.
Internet Court ร GenLayer
Slow. Expensive. Wrong.
It's been this way for centuries. It doesn't have to be.
#JusticeNeedsAnUpgrade
fatty retweeted
๐บ๐ฒ: โ๐ ๐๐๐โฆ ๐๐๐ ๐ข๐๐ ๐๐๐๐๐๐๐ ๐๐ ๐ ๐๐๐ ๐๐๐ ๐๐ก๐๐๐๐๐๐ ๐๐ ๐๐๐ ๐.๐. ๐๐๐๐๐๐๐ ๐๐๐๐๐๐ ๐๐๐๐๐๐ ๐ ๐๐๐๐๐๐ ๐๐๐ข๐๐๐ ๐ ๐๐๐๐๐?โ
๐๐ผ๐: โ๐ข๐๐, ๐๐๐ ๐ข๐๐ ๐๐๐โ๐ ๐๐๐๐ ๐๐๐๐๐๐๐๐๐ ๐๐ ๐๐๐๐๐๐๐ ๐๐ ๐๐๐ ๐๐๐๐๐๐๐.โ
๐บ๐ฒ: โ๐๐๐๐ ๐๐๐ ๐๐๐๐ ๐๐๐ ๐ ๐ ๐๐๐๐๐๐๐ ๐๐๐๐๐?โ
๐๐ผ๐: โ๐๐๐๐ ๐ท ๐๐๐๐ ๐๐๐๐๐๐๐ ๐ฝ๐๐
๐ฐ ๐ท๐พ๐ผ๐ด.โ
๐บ๐ฒ: โ๐ต๐ผ๐น๐ฑ ๐ผ๐ปโฆ ๐ฟ๐ฒ๐ฎ๐น ๐ฒ๐๐๐ฎ๐๐ฒ ๐ฐ๐ฟ๐ฒ๐ฑ๐ถ๐, ๐ผ๐ป๐ฐ๐ต๐ฎ๐ถ๐ป, ๐๐๐ฎ๐ฟ๐๐ถ๐ป๐ด ๐ณ๐ฟ๐ผ๐บ ๐ญ ๐๐๐ฑ๐ฐ? ๐ป๐ผ๐ ๐๐ผ๐ ๐ต๐ฎ๐๐ฒ ๐บ๐ ๐ฎ๐๐๐ฒ๐ป๐๐ถ๐ผ๐ป.๐โ
i was navigating the rwas space, looking for projects doing something meaningful with real-world assets, when i came across @NUVAFinance.
and honestly, this got me thinking.
what if accessing opportunities connected to the u.s. housing market didnโt require you to buy property or have substantial capital?
๐ก HOME is live by NUVA Finance : US. housing credit, onchain, from 1 usdc.
hereโs why iโm paying attention.
๐ฅ๐๐๐ ๐๐ฆ๐ง๐๐ง๐ ๐๐ฆ ๐ ๐ข๐ฅ๐ ๐ง๐๐๐ก ๐ข๐ช๐ก๐๐ก๐ ๐ฃ๐ฅ๐ข๐ฃ๐๐ฅ๐ง๐ฌ
when people talk about real estate, they usually think about buying land, building houses or purchasing apartments to rent out.
but thereโs another side to the market: credit.
in the United States, homeowners can borrow against their home equity through ๐๐ผ๐บ๐ฒ ๐๐พ๐๐ถ๐๐ ๐๐ถ๐ป๐ฒ๐ ๐ผ๐ณ ๐๐ฟ๐ฒ๐ฑ๐ถ๐ (๐๐๐๐ข๐๐ฆ).
these loans generate interest, creating a potential income source for investors exposed to the underlying credit.
๐๐ข๐ ๐ brings exposure to a managed portfolio of these u.s. residential housing loans into an onchain investment structure.
๐ช๐๐ฌ ๐ฆ๐ง๐๐ฅ๐ง๐๐ก๐ ๐ช๐๐ง๐ ๐ญ ๐จ๐ฆ๐๐ ๐ ๐๐ง๐ง๐๐ฅ๐ฆ
traditional private credit opportunities can require substantial capital and connections that smaller investors may not have.
HOME introduces a different entry point.
โช๏ธ ๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐: ๐ท ๐๐๐๐
โช๏ธ ๐๐๐๐ ๐๐๐: ๐๐๐๐๐๐๐๐
โช๏ธ ๐๐๐๐๐๐๐ ๐๐๐๐๐ ๐๐๐: $๐ท๐ธ ๐๐๐๐๐๐๐
โช๏ธ ๐๐๐๐๐๐ ๐๐๐: ๐๐๐๐๐๐ก๐๐๐๐๐๐๐ข $๐ท๐ถ ๐๐๐๐๐๐๐
โช๏ธ ๐๐๐๐๐๐๐ข๐๐๐ ๐๐ก๐๐๐๐๐๐: ๐.๐. ๐๐๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐ ๐๐๐๐๐๐.
you deposit USDC into the HOME vault and receive ๐ป๐๐๐ข๐ ๐ tokens representing your share of the vault.
your 1 usdc doesnโt buy a house or give you ownership of someoneโs property.
it gives you exposure to a managed housing-credit strategy through a tokenized structure.
thatโs the interesting part: accessing real-world credit without needing the capital to purchase real estate yourself.
๐ ๐ณ.๐ฌ% ๐ง๐๐ฅ๐๐๐ง ๐ฅ๐๐ง๐ ๐ช๐๐ง๐ ๐ ๐๐๐ฅ๐ฆ๐ง-๐๐ข๐ฆ๐ฆ ๐๐จ๐๐๐๐ฅ
HOME targets a 7.0% annual policy rate, reviewed monthly, with HELOCs at launch.
but thereโs more to consider than the headline yield.
โช๏ธ a first-loss buffer targets 5% of the vault cap, designed to absorb portfolio losses before they reach nvHOME holders.
โช๏ธ Figure Investment Advisors LLC, an SEC-registered investment adviser, manages the portfolio.
โช๏ธ a 5% liquidity sleeve supports redemptions, typically processed within three U.S. business days.
these features are designed to support risk management and liquidity, but they donโt eliminate investment risk.
the 7.0% is a target, not a guaranteed return.
๐ช๐๐ฌ ๐ง๐๐๐ฆ ๐ ๐๐ง๐ง๐๐ฅ๐ฆ ๐๐ข๐ฅ ๐ฅ๐ช๐๐ฆ ๐๐ก๐ ๐๐๐๐
what interests me most isnโt simply putting housing credit on a blockchain.
itโs connecting traditional financial assets with infrastructure that crypto users can access.
1, ๐๐ถ๐ด๐๐ฟ๐ฒ ๐ง๐ฒ๐ฐ๐ต๐ป๐ผ๐น๐ผ๐ด๐ ๐ฆ๐ผ๐น๐๐๐ถ๐ผ๐ป๐ (๐ก๐๐ฆ๐๐๐ค: ๐๐๐๐ฅ), working with more than 600 lending partners, originates the loans,
2, while Figure Investment Advisors manages the portfolio.
nvHOME is also designed to integrate with supported defi applications, creating possibilities for its use beyond the initial deposit.
๐๐ถ๐๐ต ๐ฎ $๐ญ๐ฎ ๐บ๐ถ๐น๐น๐ถ๐ผ๐ป ๐ถ๐ป๐ถ๐๐ถ๐ฎ๐น ๐๐ฎ๐๐น๐ ๐ฐ๐ฎ๐ฝ, ๐๐ข๐ ๐ ๐ต๐ฎ๐ ๐ฎ ๐ฑ๐ฒ๐ณ๐ถ๐ป๐ฒ๐ฑ ๐๐๐ฎ๐ฟ๐๐ถ๐ป๐ด ๐ฐ๐ฎ๐ฝ๐ฎ๐ฐ๐ถ๐๐ ๐ฟ๐ฎ๐๐ต๐ฒ๐ฟ ๐๐ต๐ฎ๐ป ๐ฎ๐ป ๐๐ป๐น๐ถ๐บ๐ถ๐๐ฒ๐ฑ ๐ฝ๐ผ๐ผ๐น.
my take?
the future of rwas shouldnโt just be about turning assets into tokens.
it should also be about making the financial opportunities behind those assets more accessible.
HOME is an interesting example of how traditional housing credit can meet onchain finance, with a low entry point and a structure worth understanding.
of course, low entry requirements donโt mean low risk. research the terms before participating.
1 usdc. US housing credit. onchain access.
thatโs the idea behind NUVA HOME. ๐ก
deposit with 1 usdc & explore: app.nuva.finance/
1/ Introducing HOME, NUVAโs flagship vault.
The first ever permissionless DeFi vault that directly holds a portfolio of US mortgage loans.
Monthly fixed policy rate targeting 7.00%, supported by a first-loss buffer.
Loans originated by @Figure and partners.
Live now.
fatty retweeted
AI agents are becoming capable of doing more than answering questions. They can carry out tasks, make payments, and work with other agents.
But there is one part of this that deserves more attention.. what happens when a deal goes wrong?
A company hires an AI agent to find the best supplier for its business. The agent delivers its findings, but the company discovers that some of the information is incorrect.
The company believes the job was poorly done. The agentโs side says the task was completed.
Now there is a disagreement, and someone needs to decide what happens next.
Sending payments wonโt settle it. Knowing who the two parties are wonโt settle it either. Neither will simply allowing them to communicate.
Someone needs to review the case and decide which side has the stronger argument.
This is one of the things @GenLayer is building towards.
GenLayer is a blockchain that brings AI-based judgment into smart contracts. This means contracts can handle certain decisions that are difficult to settle using fixed rules alone.
One of its key developments is Internet Court. It launched with 27 founding members and has since added Solana, Yellow, HOL and Blocksee.
When a dispute is reviewed, validators run different AI models to assess the available evidence. They then work towards an agreed verdict.
I find this important because business agreements donโt always go as planned. As AI agents take on more work, there will be situations where the outcome isnโt clear and both sides believe they are right.
Having a way to review those cases could help agents do business with less need for people to step in every time a disagreement occurs.
For me, thatโs what makes GenLayer worth following. It is exploring how smart contracts can handle decisions that go beyond simply following instructions.
As AI agents take on more responsibility, would you trust different AI models to help settle a dispute, or would you prefer a human to make the final decision?
fatty retweeted
What happens when an AI agent says it has done the job, but the other side disagrees?
Imagine a business pays an AI agent to study the market before launching a new product.
The agent sends its report, but the business finds some wrong information in it. the agent says it has done the job, while the business feels it didnโt get what it paid for.
Both sides believe they are right. so, who decides what happens next?
AI agents can already send payments, identify who they are dealing with, and communicate with other systems.
But these things donโt help much when two sides disagree about whether a job was done properly.
The business might lose its money, or a person might have to step in and settle the matter. this makes it harder for AI agents to handle business on their own.
@GenLayer is working on a way to handle these disagreements.
GenLayer uses different AI models to review the available information and agree on a decision. this helps smart contracts handle disagreements that ordinary computer rules may not be able to settle.
As AI agents take on more tasks and handle more payments, they will need a fair way to settle disputes when things go wrong.
Being able to make a deal is one thing. having a way to settle disagreements is just as important.
The internet never had a judge.
AI agents are about to move $9 trillion. Nobody built the part that handles disagreement.
GenLayer did.
And now you can become part of it: portal.genlayer.foundation/gโฆ
AI agents are becoming capable of doing more than answering questions. They can carry out tasks, make payments, and work with other agents.
But there is one part of this that deserves more attention.. what happens when a deal goes wrong?
A company hires an AI agent to find the best supplier for its business. The agent delivers its findings, but the company discovers that some of the information is incorrect.
The company believes the job was poorly done. The agentโs side says the task was completed.
Now there is a disagreement, and someone needs to decide what happens next.
Sending payments wonโt settle it. Knowing who the two parties are wonโt settle it either. Neither will simply allowing them to communicate.
Someone needs to review the case and decide which side has the stronger argument.
This is one of the things @GenLayer is building towards.
GenLayer is a blockchain that brings AI-based judgment into smart contracts. This means contracts can handle certain decisions that are difficult to settle using fixed rules alone.
One of its key developments is Internet Court. It launched with 27 founding members and has since added Solana, Yellow, HOL and Blocksee.
When a dispute is reviewed, validators run different AI models to assess the available evidence. They then work towards an agreed verdict.
I find this important because business agreements donโt always go as planned. As AI agents take on more work, there will be situations where the outcome isnโt clear and both sides believe they are right.
Having a way to review those cases could help agents do business with less need for people to step in every time a disagreement occurs.
For me, thatโs what makes GenLayer worth following. It is exploring how smart contracts can handle decisions that go beyond simply following instructions.
As AI agents take on more responsibility, would you trust different AI models to help settle a dispute, or would you prefer a human to make the final decision?
fatty retweeted
the fomo on @axisrobotics is incredibly crazy.
in less than one week, more than 500 persons have joined the kaito epoch 2 campaign.
didnโt expect this amount of turn up in the very last hours.
wishing everyone goodluck, tuff.
update on @axisrobotics
you have until October 18, 2026, at 12:00 UTC to sign your records/trajectories on Axis.
this will most likely be the final EPOCH before the $AXIS TGE.
failure to sign your trajectories will result in your points and airdrop being transferred to the 6% pool for the post-TGE airdrop.
also, if youโre participating in the Kaito creators campaign, you have just 3 days left to submit your content on Kaito
as this will be the final epoch before TGE as well.
good luck and gAXIS to those who believe in the future of embodied AI.
What happens when an AI agent says it has done the job, but the other side disagrees?
Imagine a business pays an AI agent to study the market before launching a new product.
The agent sends its report, but the business finds some wrong information in it. the agent says it has done the job, while the business feels it didnโt get what it paid for.
Both sides believe they are right. so, who decides what happens next?
AI agents can already send payments, identify who they are dealing with, and communicate with other systems.
But these things donโt help much when two sides disagree about whether a job was done properly.
The business might lose its money, or a person might have to step in and settle the matter. this makes it harder for AI agents to handle business on their own.
@GenLayer is working on a way to handle these disagreements.
GenLayer uses different AI models to review the available information and agree on a decision. this helps smart contracts handle disagreements that ordinary computer rules may not be able to settle.
As AI agents take on more tasks and handle more payments, they will need a fair way to settle disputes when things go wrong.
Being able to make a deal is one thing. having a way to settle disagreements is just as important.
The internet never had a judge.
AI agents are about to move $9 trillion. Nobody built the part that handles disagreement.
GenLayer did.
And now you can become part of it: portal.genlayer.foundation/gโฆ
fatty retweeted
a quick recap of yesterdayโs @RialoHQ builders hub.
two projects were presented during the session, each with a different idea for improving how people use blockchain applications.
hereโs a quick breakdown of what was shared.
โ ๐๐ฟ๐ฎ๐๐ฏ๐ผ๐ผ๐ธ
drawbook focuses on making onchain raffles fair and verifiable. users buy tickets and commit a secret hash linked to their wallet. when the draw takes place, the contract verifies the entries and selects a winner in a single transaction.
the idea is to make the results transparent and reduce the possibility of manipulation.
โ ๐ฅ๐ถ๐ฎ๐น๐ผ ๐๐ฎ๐น๐น๐ฒ๐
the second project focused on making digital asset management easier.
rialo wallet brings essential functions into one place, including a unified asset dashboard, options to send and receive assets, and an activity ledger to track transactions.
it gives users a more direct way to manage their assets without having to move between different tools for basic operations.
both presentations showed how builders are working on practical problems, from making raffle results verifiable to making everyday blockchain interactions easier to manage.
gRialo. see you at the next builders hub on wednesday
