Document the live research prototype

README.md

@@ -18,7 +18,7 @@ tags:
 
 ### Open intelligence for the commons.
 
-**Status: concept project. No model weights or hosted inference endpoint are included yet.**
+**Status: working public-data prototype. No model weights or generative inference endpoint are included yet.**
 </div>
 
 Robinhood is an open AI research concept for making public information easier to understand. Its intended role is to gather evidence, explain uncertainty, and help people learn how markets and public systems work without pretending to be a financial adviser.
@@ -27,6 +27,12 @@ The name draws from the public-domain Robin Hood legend: knowledge should not re
 
 **Official website:** [Robinhood-Web](https://huggingface.co/spaces/pumpfun/Robinhood-Web)
 
+## Live research prototype
+
+The official website can now resolve a U.S. public company by ticker or name and retrieve live SEC EDGAR submissions plus standardized XBRL company facts. It presents recent material filings, selected reported financial facts, their periods, and direct links to the original SEC documents.
+
+This first functional release is an evidence-retrieval system rather than a trained language model. Its output is deterministic, and it does not create recommendations or predictions.
+
 ## The idea
 
 Robinhood is designed around four promises: