feat: add service loader + SPEC_06 + P0 bug report

fix: implement SPEC_04 (Magentic UX) and SPEC_05 (cleanup)

feat: Implement SPEC_01 (Termination) and SPEC_02 (E2E Tests)

fix: P0/P1/P2 - Gradio crash, UX improvements, thinking state (#61)

Merge branch 'dev' - P1 bug fixes + CodeRabbit feedback

fix: replace deprecated OpenAIModel with OpenAIChatModel

fix(config): use gpt-5.1 (actually available) not gpt-5 (deprecated)

fix: P0 bugs + model config stabilization (November 2025) (#59)

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

fix: complete audit fixes for documentation accuracy

rebrand: DeepCritical → DeepBoner (sexual health research agent)

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

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

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

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

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

chore: clean up env config and upgrade model defaults

fix: make ResearchReport.references optional with default

refactor: address CodeRabbit nitpick suggestions

feat: implement phase 13 modal pipeline integration

feat: Phase 11 bioRxiv Preprint Integration (#19)

refactor: centralize embedding configuration in Settings

feat: Integrate Mario's Modal code execution + LlamaIndex RAG

refactor: address CodeRabbit nitpicks for Phase 10

feat: Implement Phase 10 (ClinicalTrials.gov) with requests

feat: Complete Phase 9 Source Cleanup (Remove WebTool) (#14)

fix: address all CodeRabbit Phase 8 review feedback

fix: apply lazy init pattern and pydantic-ai fixes to ReportAgent (#12)

fix: address CodeRabbit Phase 7 review feedback

feat(phase7): implement Hypothesis Agent for mechanistic reasoning

fix: address CodeRabbit Phase 5 review feedback

fix: linting and types for phase 5

fix: pass API keys explicitly to pydantic-ai providers

fix: pydantic-ai v1 compatibility and Anthropic model update

refactor(phase4): address CodeRabbit review

feat(phase4): Orchestrator and UI complete

feat(phase3): implement judge slice (LLM Judge, Prompts, Models)

feat(phase2): implement search slice (PubMed, Web, Orchestrator) (#4)

feat(phase1): Foundation & Tooling complete

adds scaffolding