What problem does it solve? Getting rigorous, senior-level feedback on research ideas, papers, and experimental results is slow and often unavailable on demand. This Skill orchestrates a deep, multi-round critical review of your research from an external OpenAI model (via Codex MCP) with maximum reasoning effort, producing actionable experiment plans and paper guidance. ## Core Features & Use Cases - Multi-Round External Review: Sends comprehensive research context to GPT via Codex MCP with xhigh reasoning, then iterates through follow-up rounds using threadId-based replies. - Actionable Deliverables: Requests experiment designs, paper outlines, mock NeurIPS/ICML reviews, and results-to-claims matrices from the reviewer. - Documented Conclusions: Saves round-by-round criticisms, final consensus, claims matrices, and prioritized TODO lists with compute estimates to a self-contained review document. - Use Case: You have a draft paper and experimental results for a video codec project. Run this Skill to get a brutal senior-reviewer critique, defend or revise your claims over several rounds, and finish with a concrete experiment plan and paper outline. ## Quick Start Ask the assistant to review my research using the research-review skill, providing the topic or scope of the work to critique.
