research-brainstorm

Coordinate Gemini, Codex, and Claude to generate and synthesize research ideas.

13|2|Updated Feb 18, 2026
One-click install
npx skills add https://github.com/Axect/magi-researchers --skill research-brainstorm-axect
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: research-brainstorm
Source: https://github.com/Axect/magi-researchers/tree/main/skills/research-brainstorm
Command: npx skills add https://github.com/Axect/magi-researchers --skill research-brainstorm-axect

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Generates and cross-validates research ideas using multiple AI agents (Gemini and Codex) in parallel, then synthesizes results with Claude to produce high-quality, publication-ready insights.

Core Features & Use Cases

  • Parallel ideation: Gemini and Codex generate diverse ideas simultaneously.
  • Cross-review: automated evaluation and feedback rounds to refine concepts.
  • Synthesis: Claude consolidates findings into structured summaries and actionable plans.
  • Output organization: artifacts are saved under outputs/{topic}/ with a reproducible naming scheme for later review.

Quick Start

Provide a topic and optional domain flags to start a full brainstorm cycle and retrieve a synthesized research plan.

Frequently Asked Questions about research-brainstorm

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

FAQPage Schema
How do I cross-validate research ideas using multiple AI agents for brainstorming?

Multi-agent brainstorming cross-validates research ideas by running parallel ideation across Gemini and Codex, then synthesizing the results with Claude to produce a cohesive, publication-ready report.

Can I generate publication-ready research insights by combining outputs from Gemini, Codex, and Claude?

Yes, you can generate publication-ready insights by using Claude-based synthesis to consolidate diverse ideas and automated feedback rounds from Gemini and Codex into structured summaries and actionable plans.

How does multi-agent synthesis work for cross-verifying academic research topics?

Multi-agent synthesis works by deploying Gemini and Codex to generate diverse ideas simultaneously, followed by automated cross-review rounds to refine concepts before Claude consolidates the validated findings.

What is the best way to organize reproducible research outputs from a multi-agent brainstorming session?

Organize reproducible research outputs by saving artifacts under a topic-specific directory with a reproducible naming scheme, enabling structured review and end-to-end publication-ready workflows.

Do I need to provide specific domain templates to start a multi-agent research brainstorming cycle?

No, you only need to provide a research topic to start a full brainstorm cycle, though optional domain flags and depth controls are supported to tailor the multi-agent ideation and synthesis process.

Are there limitations to using automated cross-review for research idea generation?

Automated cross-review relies on parallel ideation from Gemini and Codex, meaning limitations depend on the depth controls applied and the domain templates provided for the specific research topic.