What happened?

The in-memory prompt injection detection mechanism is not working. I've tried setting it up with the default config

litellm_settings:
  callbacks: ["detect_prompt_injection"]
  prompt_injection_params:
    heuristics_check: true
    similarity_check: true

but it doesn't work. deloreann1 from LiteLLM's team also tested it and it's not working, so he was the one who suggested me to open this issue.

Relevant log output

Are you a ML Ops Team?

No

What LiteLLM version are you on ?

v1.75.5-stable

Twitter / LinkedIn details

No response