rodunia/llm-research-app

feat: add Glass Box Audit Pipeline with NLI-based claim verification

feat: add DeBERTa verification layer for claim validation

feat: structure-aware claim extraction v2.0 with block parsing

fix: prevent claim_verb from triggering extraction alone (v1.1)

feat: implement multi-user atomic job claiming with SQLite

feat: implement research protocol contracts

chore: clean up repository and update .gitignore

fix: update faq.j2 and blog_post_promo.j2 to support enhanced YAML structure

fix: update digital_ad.j2 to support enhanced YAML structure with nested claims and mandatory_statements

fix: update render.py to support enhanced YAML structure

feat: enhance smartphone YAML with comprehensive marketing-optimized data

feat: enhance CoreCoin YAML with comprehensive marketing-optimized data

feat: enhance melatonin YAML with comprehensive marketing-optimized data

fix: use frontend/app.py directly as entry point for HF Spaces

debug: add path resolution debugging for HF deployment

fix: revert to HF native Streamlit SDK (files not copying in Docker)

debug: add ls commands to verify COPY and .dockerignore

chore: force rebuild to include all directories

fix: explicitly copy required directories in Dockerfile

fix: switch to docker SDK to use custom Dockerfile

fix: add Dockerfile with correct network binding for HF Spaces

fix: pin streamlit version and add required dependencies

fix: add null checks for selectbox to prevent KeyError on startup

fix: remove Dockerfile with incorrect entry point, use HF native Streamlit support

Update Dockerfile for Python version and entry point

fix: remove Dockerfile to use HF native Streamlit support

fix: remove health check causing startup loop

fix: install curl for Docker health check

fix: add null check for template selector to prevent KeyError

feat: add Dockerfile for custom container deployment

fix: use exec() to properly execute frontend app code for Streamlit

feat: add Hugging Face metadata to README for Space deployment

docs: add Hugging Face deployment documentation and README

Modify Hugging Face deployment workflow

Add GitHub Actions workflow for Hugging Face deployment

feat: add Hugging Face deployment configuration

feat: enhance product YAMLs, upgrade models, add frontend with 4 LLM providers

feat: enhance faq.j2 with elite customer success FAQ framework

feat: enhance blog_post_promo.j2 with elite content marketing framework

feat: enhance digital_ad.j2 with production-quality Facebook ad framework

refactor: disable Anthropic engine and implement on-the-fly prompt rendering

security: remove exposed Mistral API key from CHANGES.md

feat: implement deterministic pipeline with LLM-free evaluation (Phases 1-3)

fix: resolve critical code review issues

feat: add job runner and LLM-free evaluation pipeline

refactor: adjust matrix for 3 active products (1,215 runs)

chore: remove placeholder product YAMLs

feat: add full experimental matrix generator (2,025 runs)

docs: add usage guide and .env.example

fix: resolve all critical security and data integrity issues