Re:Supabase MCP can leak an entire SQL Database, a lethal tr...
by ImPostingOnHN on Tuesday July 08, 2025 @11:03PM
(#44506007)
Why would it be strange or harder?
The initial prompt can contain as many layers of inception-style contrivance, directed at as many inaginary AI "roles", as the attacker wants.
It wouldn't necessarily be harder, it'd just be a prompt that the attacker submits to every AI they find.
Both layers failing isn’t impossible, but it’d be much harder than defeating the existing protections.