refactor(judges): extract shared title-extraction helper

fix(config): update default LLM models to stable Nov 2025 versions

fix(judges): fallback to heuristic extraction when HF quota exhausted

fix: resolve all P0 critical bugs blocking demo

fix(ui): clarify mode compatibility in radio button info

fix(ui): apply CodeRabbit review suggestions

docs: add UI/UX brainstorming for mode selection

fix(ui): remove empty API key column from examples table

fix(ui): simplify BYOK settings with auto-detection

fix: remove css param not supported by ChatInterface

style: center header text in Gradio UI

fix: complete audit fixes for documentation accuracy

fix: address CodeRabbit review feedback

fix: resolve all test warnings and complete DeepBoner rebrand

rebrand: DeepCritical → DeepBoner (sexual health research agent)

style: auto-format 4 files (ruff)

feat: add hierarchical orchestration with web search (integrates PR #44) (#46)

feat: implement dual-mode architecture (Simple + Advanced) (#45)

feat: implement async-safe rate limiting (Phase 17) (#40)

feat: add roadmap summary and detailed improvement plans for data sources

fix: address CodeRabbit feedback

fix: repair Gradio accordion bug and cleanup env vars

fix: address CodeRabbit feedback and CI failure

fix(magentic): defensive text extraction & cleanup biorxiv refs

fix: update references to Europe PMC in examples and documentation

fix(config): increase max iterations to 10

feat(search): implement ClinicalTrials.gov filtering (Phase 03)

feat(search): implement PubMed query preprocessing (Phase 02)

refactor(tools): replace BioRxiv with Europe PMC (Phase 01)

fix: exclude reference_repos from mypy and cleanup

refactor: enhance Gradio UI for improved user experience

fix: simplify Gradio UI to fit single viewport

fix: address CodeRabbit feedback and P0 blockers

docs: add P0 blockers documentation for Magentic mode implementation

feat: implement Magentic ChatAgent pattern with semantic state management

feat: implement HFInferenceJudgeHandler for free-tier AI analysis

fix: wire EmbeddingService to simple orchestrator + improve search quality

fix: disable ssr_mode to resolve intermittent gradio hydration issues

fix: resolve Gradio UI cutoff by removing fill_height=True and restructuring layout

fix: correct UI layout ordering - title at top

fix: move title/description into ChatInterface for proper HF Spaces layout

fix: upgrade Gradio SDK to 6.0.1 and add fill_height for HF Spaces UI fix

fix: address CodeRabbit review feedback

chore: clean up env config and upgrade model defaults

fix: use config settings for model names instead of hardcoded values

feat: add demo mode warnings and BYOK (Bring Your Own Key) support

fix: upgrade Gradio to 5.32.0+ for MCP server support

feat: add smoke tests and fix Gradio 6.x compatibility

fix: Gradio examples format for additional_inputs

fix: make ResearchReport.references optional with default