small cleanups

Included latest version of pysqlite to the requirements.

Small cleanup

Added userId and date-time to the conversations.

Final application code deployment. Ensures all functionalities, including correct SYSTEM_PROMPT usage, Firestore conversation saving, and returning conversation ID in API response.

Fix: 'HEADERS' not defined error by using DEEPSEEK_HEADERS from config.py in compassia.py.

Small updates on app.py

Small updates on app.py

CRITICAL FIX: Ensured pysqlite3 override is the absolute first action in app.py for ChromaDB compatibility on HF Spaces.

Edited sqlite3 to ensure its accessibility.

Fix: ChromaDB SQLite version issue for Hugging Face deployment by moving pysqlite3 import to app.py. Refined doc indexing.

Small updates on the app file.

Reorganized the project: -Documents hosted on Firestore db, -conversations saved

Fix: Move pdf_document_paths to global scope in compassia.py for import

Fix: Move pdf_document_paths to global scope in compassia.py for import

Fix: Add .local/bin to PATH for uvicorn and other installed executables

Fix: Change ownership of /app to user in Dockerfile

Fix: Run apt-get as root before switching to user in Dockerfile

Fix: Set HF_HOME explicitly and add cache cleanup for model download permissions

Added prompt system to compassia

Fix: Force pysqlite3 usage for ChromaDB compatibility

Fix: Use pysqlite3-binary for ChromaDB compatibility and simplify Dockerfile

Adjust app.py to import from compassia.py in root

Final fix: Updated requirements.txt

Final fix: Ensure apt_packages line is removed from requirements.txt

Modified requirements.txt file.

Metadata repo card added - README.md

Configure Git LFS to track PDFs and ChromaDB SQLite

Ensure chroma_db is untracked

Stop tracking chroma_db and add it to .gitignore

Add FastAPI app and Dockerfile