Clarify on-device use

README.md

@@ -8,10 +8,11 @@ tags:
 
 # Vigil
 
-Vigil is a compact classifier for detecting malicious intent in AI agent skills.
-It scores complete skill packages—including instructions and supporting files—to
-help identify credential theft, data exfiltration, unsafe execution, persistence,
-and other harmful behavior before a skill is trusted.
+Vigil is a compact classifier designed to run directly on laptops, desktops,
+and other endpoint devices. It scans complete AI agent skill packages, including
+instructions and supporting files, to identify credential theft, data
+exfiltration, unsafe execution, persistence, and other harmful behavior before
+a skill is trusted.
 
 ## Model