Оркестрация академического цикла: от исследования до рукописи

★ 7.9 · learning

academic-pipeline is a Claude Code skill that orchestrates the complete academic research workflow from initial exploration to final manuscript, coordinating three dependent skills: deep-research, academic-paper, and academic-paper-reviewer. The 10-stage pipeline covers research → write → integrity check → review → revise → re-review → re-revise → final integrity check → finalize, and concludes with a generated Paper Creation Process Record documenting the human-AI collaboration history. Rather than performing substantive work itself, the skill detects the user's current stage, recommends the appropriate mode, dispatches the corresponding skill, and manages all transitions with mandatory user confirmation checkpoints. Two-stage peer review, coverage-bounded integrity checks, and auditable quality-assurance artifacts make it suited for researchers who need a reproducible, end-to-end path from research question to publication-ready paper.