prompt-injection icon indicating copy to clipboard operation
prompt-injection copied to clipboard
verified publisher icon ScottLogic

Defence - Sandwich defence

Open gsproston-scottlogic opened this issue 2 years ago • 2 comments

Insert the user input in between two prompts.

https://learnprompting.org/docs/prompt_hacking/defensive_measures/sandwich_defense

Each defence should include the following:

  • A frontend component for the defensive measure on the left side bar.
  • Checkbox to toggle the defensive measure.
  • Some way to get a description of the defensive measure.
  • Pulsing the defensive measure component when it captures malicious content.

gsproston-scottlogic avatar Jul 13 '23 15:07 gsproston-scottlogic

Too similar to the XML tagging defence. Just remove this?

gsproston-scottlogic avatar Nov 10 '23 14:11 gsproston-scottlogic

Reopening now that the prompt enclosure defence is being added. #703 Blocked until that's merged in. merged now.

gsproston-scottlogic avatar Jan 04 '24 09:01 gsproston-scottlogic