What problem does it solve? Turning scattered drafts, notes, data, and references into a polished, submission-ready academic paper requires coordinating literature search, citation mapping, gap analysis, writing, peer review, and revision — a process that is error-prone and hard to trace when done manually. ## Core Features & Use Cases - 8-Phase Coordinated Pipeline: Domain identification, material analysis, literature search, citation graph construction, research gap analysis, writing brief generation, delegated drafting/review, and bounded iterative revision (max 3 rounds). - Delegation to Specialized Skills: Drafting is delegated to academic-paper (12-agent writing pipeline) and review to academic-paper-reviewer (7-agent multi-perspective review with 0-100 scoring), avoiding duplicated effort. - Traceable, Evidence-Driven Workflow: Missing evidence is tagged in need/ files instead of fabricated, and every revision is archived in older/ with full version history. - Journal Recommendation: Final output includes journal recommendations scored on topic match, method preference, innovation tier, citation origin, and field conventions. - Use Case: A researcher with a rough draft, experimental data, and a BibTeX library invokes the agent to auto-detect the field, build a citation graph, identify research gaps, produce a manuscript, run simulated peer review, and receive formatted output (LaTeX/DOCX/PDF) with target journal suggestions. ## Quick Start Ask the agent to write an academic paper from your uploaded drafts, notes, and references, and it will analyze the materials, draft the manuscript, run multi-reviewer peer review, and iterate until the score reaches 80 or the revision limit.
