skills.rest
OpenLAIR
OpenLAIROFFICIAL
Explore all skills by this author →
GitHub

aris-research-review

Obtains multi-round critical research reviews from GPT via Codex MCP with xhigh reasoning.

1.1k|116|Updated Feb 26, 2026
Search Tags:#experiment design#gpt#peer review#research review#codex mcp#paper feedback#Machine learning
One-click install
npx skills add https://github.com/OpenLAIR/dr-claw --skill aris-research-review
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: aris-research-review
Source: https://github.com/OpenLAIR/dr-claw/tree/main/skills/aris-research-review
Command: npx skills add https://github.com/OpenLAIR/dr-claw --skill aris-research-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Getting rigorous, senior-level feedback on ML research ideas, papers, and experiments is hard without access to expert reviewers. This Skill orchestrates a multi-round critical review from an external GPT model (via Codex MCP) acting as a NeurIPS/ICML-level reviewer, producing actionable experiment plans and claims matrices.

Core Features & Use Cases

  • Deep External Review: Sends comprehensive research context to GPT (e.g., gpt-5.4) with xhigh reasoning effort to identify logical gaps, missing experiments, and narrative weaknesses.
  • Iterative Dialogue: Continues multi-round conversations via threadId to respond to criticisms, request experiment designs, mock reviews, and results-to-claims matrices.
  • Documented Outcomes: Saves round-by-round summaries, final consensus on claims, prioritized TODO lists with compute estimates, and paper outlines to a review document.
  • Use Case: Before submitting a paper, ask for an external review of your draft; the Skill compiles your project context, gets brutal reviewer feedback, iterates on rebuttals, and produces a concrete experiment plan to strengthen the submission.

Quick Start

Ask the assistant to review my research paper draft using the external GPT reviewer and produce an experiment plan addressing the criticisms.

Frequently Asked Questions about aris-research-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I get an external AI review of my research paper?▼

Compile your research context (paper drafts, experiment notes, key claims) and send it to GPT via the Codex MCP tool with xhigh reasoning effort. The reviewer identifies logical gaps, missing experiments, and narrative weaknesses, then iterates with you across multiple rounds.

How to set up Codex MCP for research review in Claude Code?▼

Run 'claude mcp add codex -s user -- codex mcp-server' to register the Codex MCP server. This exposes the mcp__codex__codex and mcp__codex__codex-reply tools used for the initial review and follow-up dialogue rounds.

Which models can act as the external reviewer via Codex MCP?▼

The reviewer must be an OpenAI model such as gpt-5.4, o3, or gpt-4o. The default is gpt-5.4, and every review call should set model_reasoning_effort to xhigh for maximum reasoning depth.

Can I continue a research review conversation across multiple rounds?▼

Yes, use the mcp__codex__codex-reply tool with the threadId returned from the initial review call. This lets you respond to criticisms, request experiment designs, mock reviews, and claims matrices until consensus is reached.

What are the limitations of LLM-based research review?▼

The external model cannot read your local files, so all relevant context must be included in the first prompt. Feedback quality also depends on honest disclosure of weaknesses, and outputs should be documented since the review is advisory rather than a substitute for peer review.

Related Skills

Explore More Education & Research Skills→
big4council-progbig4council-prog

research-paper-writing

Guide ML/AI research paper writing from project setup through submission.

Community
Advanced
Wenwen555Wenwen555

paper-plan

Extract claims and evidence from documents to generate a structured paper outline.

Community
Advanced
UnaryLabUnaryLab

literature-survey

Search academic indexes and generate a cited Markdown literature survey.

Official
Advanced
obraobra
41

Code Review Reception

Evaluate and implement code review feedback with technical verification.

Community
Basic