Loops Houselithium

NEAR HACKATHON
Links
Team
1 member- SAOwner
sammy
Overview
The problem: NEAR Intents volume is often read as a growth signal, but investors can't verify whether the solvers filling it are diversified or depend on a few operators. A static dashboard can't answer that, because the answer isn't in the data yet.
What I built: Lithium, an interactive tool that shows what each possible answer would mean. Across 555 sampled swaps, public receipts attribute 235 to 15 executing accounts, and those accounts are not verified independent operators. So instead of claiming a verdict, Lithium lets an analyst explore the possibilities: enter hypothetical operator disclosures and watch the concentration bounds move; remove top accounts and see how much sampled revenue exposure follows, with unknown attribution kept in both endpoints. It also includes a checklist of what to demand from settlement maintainers and operators, and explicit breaking points that would change the verdict. Everything rebuilds offline with make reproduce.
How NEAR fits: the evidence is NEAR's own public settlement data (Intents receipts via NEAR RPC and Explorer, plus a BigQuery event inventory), and the revenue inputs are cited from another entry's NEAR valuation work. Lithium turns an untested assumption in NEAR Intents' growth story into something testable. These are scenarios, not predictions: the sample is small (48 minutes across 8 dates), and every limitation is listed in the README.