What problem does it solve? Coordinating a full research-to-publication workflow manually means juggling literature research, drafting, citation integrity checks, peer review simulation, and revision cycles across disconnected steps, with no audit trail or quality gates between stages. ## Core Features & Use Cases - Ten-stage orchestration: Dispatches deep-research, academic-paper, and academic-paper-reviewer skills across RESEARCH, WRITE, INTEGRITY, REVIEW, REVISE, RE-REVIEW, FINAL INTEGRITY, FINALIZE, and PROCESS SUMMARY stages with mandatory user confirmation checkpoints. - Integrity verification gates: Runs coverage-bounded reference, citation, claim, and data checks plus a 7-mode AI research failure checklist at Stages 2.5 and 4.5, blocking the pipeline on suspected hallucinated citations or fabricated results. - Two-stage peer review with audit trail: Full five-reviewer review followed by post-revision verification review, with versioned artifacts, a Material Passport for cross-session resume, and a bilingual process record PDF. - Use Case: A graduate student starts with only a topic, receives a researched and drafted paper, gets simulated reviewer feedback, revises through two rounds, and finishes with a formatted manuscript plus a documented human-AI collaboration record. ## Quick Start Tell the assistant you want to write a research paper on your topic from scratch and it will launch the pipeline at Stage 1, guiding you through each stage with confirmation checkpoints.
