interactive code and security concern

[mhd wrk]

Considering that Zeppelin (via Livy) supports Interactive Scala code, does
it provide any sand-boxing feature to protect the back-end against
malicious code?