Added SBERT semantic routing and EDA Safety Guard Rails

Fixed some minor bugs

Added Query Awareness , Success/Failure Marking further improving pipeline accuracy

Fixed File path not found issues and Multi-user issues

Made Major Changes to the Overall User Experience by adding a pipeline visualizer

Fixed Bugs where the SSE Streaming was improper added support for Auto Gluon, Fixed instances where stale schemas was causing EDA plots generation to fail

Fixed all output path issues

Align HuggingFace API usage with official docs pattern

Add better error logging to HF export endpoint

Add better error logging to HF export endpoint

Fix HuggingFace export: correct method signatures and add tmp directory

Fix HuggingFace export: correct class name and constructor

Fixed the Huggingface Hub Error

Directly making API calls to the server -V3

Directly making API calls to the server -V2

Fixing the issues with the SSE streaming all the history

Fixed the SSE Streaming not working

Fixed the Agent Routing Logic which was causing issues

Added the functionality for users to save their generated assets in huggingface spaces using write tokens

Add HuggingFace storage integration - users can now persist datasets, models, and plots to their own HuggingFace account

Make system prompt examples generic - remove earthquake-specific hardcoded references

Fix infinite thinking when no file uploaded - properly handle analysis_failed SSE event

Fix Supabase auth on HuggingFace with runtime config injection

CRITICAL: Fixed race conditions, session cleanup, SSE leaks, and added localStorage persistence

Fixed multi-session support and follow-up queries - Session UUID reuse, multi-chat isolation, proper SSE switching

Fixed slow Agent loading which caused the Agent took too long to respond

Fixed Scalability issues

Updated the Readme

Fixed the Agent giving stale responses

Fix:Auto Pruning

Fix:Auto Propping

Fix: Better SBERT error handling + suppress invalid hand-off warnings

Fixed bugs 2

Fixed bugs

Fixed known model training issues

Fixed JSON Serializability error

Fixed conversation pruning logic

Fixed the broken Attributes of Parallel executor

Fixed the Pydantic Model Errors for the Token Budget

Fixed the File Path Issues

Fixed Agent Reasoning and Semanttic Layer Integration issues

Fixed Agent Reasoning-*1

Fixed Agent Reasoning-1

Fixed Agent Reasoning

Fixed the Mistral SDK Version mismatch and improved caching efficiency , Inter Agent Communication

feat: Add 4 major system improvements - semantic layer, error recovery, token budget, parallel execution

Remove aggressive text cleanup that was breaking markdown formatting

Show LLM reasoning in summary instead of generic template

Fix JSON serialization error for numpy types in SSE stream

Fix: Check for both output_path and report_path when extracting reports