Control running lakera prompt checks - pre api call OR in parallel

[krrishdholakia]

Title

Control running lakera prompt checks - pre api call OR in parallel

litellm_settings:
  guardrails:
    - prompt_injection:  # your custom name for guardrail
        callbacks: [lakera_prompt_injection] # litellm callbacks to use
        default_on: true
        callback_args: {"lakera_prompt_injection": {"moderation_check": "pre_call"}} 

Relevant issues

Fixes issue where user couldn't control how lakera calls were running

Type

🆕 New Feature

Changes

• adds check in lakera_ai.py for pre_call or in_parallel
• adds support for passing callback-specific args in guardrail config

[REQUIRED] Testing - Attach a screenshot of any new tests passing locall

If UI changes, send a screenshot/GIF of working UI fixes

• new test added to test_lakerai_prompt_injection.py

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
litellm	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Aug 6, 2024 9:49pm