passing extra data dynamically to spark job

[mhd wrk]

What's the best way in zeppelin to pass extra data (e.g. stored in current
user's session) to spark job?