There are two answers to the question "what makes an agent creditworthy?" @agenticscredit gives one. A new framework from Decentralised News gives the other. They barely overlap.
Agentics asks: how well did you trade? The ACS is built from realised performance: profitability, drawdown, consistency, longevity, win rate. The wallet's track record is the borrower.
The DN framework asks a different question: who are you, and can anyone make you pay? Its Credit Identity Bundle is ten things: the agent, an accountable principal, the current behavioural version, verified revenue, repayment history, liabilities, reputation, collateral, financial permissions, and recovery controls. Agentics measures roughly one of them: the track record.
The gap is the point. A score tells you how an agent trades. It doesn't tell you who is accountable when the agent defaults, what can be seized, or who can be pursued. The whitepaper is honest about this in its own way: "the borrower is code."
Two honest caveats. DN admits its framework has never been validated against realised default data, because no mature market for agent credit exists yet. And Agentics' answer is at least measurable today.
But the question stands. Credit is not just a prediction about performance. It's a claim on someone when the prediction is wrong. Until an agent has an accountable principal behind it, the score rates a wallet, not a borrower.
A score tells you how an agent trades. It doesn't tell you who pays when it doesn't.