The noise wasn’t random. It was feedback. During testnet, we saw a flood of packages, pull requests, accounts and transactions. Some of it was genuine participation. Some of it was spam. But the more important lesson was that people were responding to the system we had put in front of them. If volume looked like contribution, volume became the strategy. That same problem is now playing out across open source at a much larger scale. Automated pull requests, spam packages and machine-generated activity are growing faster than maintainers can meaningfully evaluate them. So we went back to the application layer. Not to change the mission, but to build a better way of distinguishing activity from contribution, and noise from signal. Identity. Provenance. Repository context. Contribution evidence. Maintainer authority. That is what the last few months have been about. We’re getting close to showing you what we’ve been building.