SafeSpaceAI / server

Commit History

  1. Update openenv test cases

    Shreya Pal
    3851cd6
  2. Update groq model

    Shreya Pal
    f7eeeee

  1. Replace algorithmic scoring with Groq contextual LLM moderation

    Shreya Pal
    62c3394
  2. Adjust moderation thresholds for stricter tiering

    Shreya Pal
    a44e3ea
  3. Refactor moderation logic to use local scoring system

    Shreya Pal
    7dca781
  4. Fix hf_moderate error handling & loading fallback

    Shreya Pal
    885d181
  5. Update deprecated Hugging Face inference pipeline endpoint to HF Router URL mapping

    Shreya Pal
    eb29ff9
  6. Extract upstream Hugging Face API errors natively dropping into UI fallback wrapper

    Shreya Pal
    84f2df9
  7. Implement robust JSON safety fallbacks blocking malformed Mistral outputs from 500 downstream crashes

    Shreya Pal
    0d79c56
  8. Swap moderation endpoint from OpenAI to free HuggingFace API

    Shreya Pal
    4bd6a99
  9. Raise transparent errors instead of passing silent ones on LLM loop bypass

    Shreya Pal
    3b43dcd
  10. Lazy load OpenAI client

    Shreya Pal
    aace552
  11. Fix OpenEnv mapping architecture matching explicit tasks

    Shreya Pal
    ada3dd4
  12. Implement two-stage AI moderation pipeline in server/app.py

    Shreya Pal
    99e1c7c
  13. Make API Key private

    Shreya Pal
    5c5b473