research-co-pilot is a Claude Code and claude.ai plugin for academic and applied research. It bundles 14 skills covering the research lifecycle — literature review, methodology design (with mandatory creative AI / ML / Big Data extensions), ethics committee stress-testing, data analysis, qualitative coding, brainstorming, manuscript drafting (with voice preservation and academic-register enforcement), replication design, grant writing, talk building, reviewer-response drafting, citation formatting, survey design, and peer review (with anchored annotations across DOCX / PDF / PPTX / LaTeX) — alongside subagents for parallel source-reading, heavy data analysis, transcript coding, manuscript drafting, and independent stats validation. Designed to refuse common AI failure modes (fabricated citations, p-hacking, themes-without-an-audit-trail, averaging away disagreement in the literature) and to support multiple languages including Hebrew.