idea-brainstorm

Generate research ideas sorted into T1–T5 novelty tiers with prior-work verification.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/UnaryLab/ai-for-research --skill idea-brainstorm-unarylab
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: idea-brainstorm
Source: https://github.com/UnaryLab/ai-for-research/tree/main/skills/idea-brainstorm
Command: npx skills add https://github.com/UnaryLab/ai-for-research --skill idea-brainstorm-unarylab

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It helps you turn a topic or open problem into a wide set of candidate research directions and then clearly separates genuinely novel ideas from incremental or likely-known ones.

Core Features & Use Cases

  • Novelty-tiered ideation: Diverges broadly and then buckets ideas into T1–T5 novelty tiers so you can quickly spot the most original directions.
  • Novelty verification before tiering: Forces each idea to be contrasted against closest prior work and what must be checked, so “novelty” is treated as a claim to verify.
  • Divergence lenses for breadth: Produces diverse angles using named lenses like cross-domain analogy, strategic ignorance (assumption reversal), constraint relaxation, scale shifts, primitive invention, bisociation, inversion, and tech speculation.
  • Optional fail-fast handoff: When you want to act, it can add an Idea Card layer with the cheapest decisive test to kill or confirm top ideas.

Quick Start

Ask the skill to brainstorm research ideas for your topic and return them as a tiered novelty list (T1 highest) with what prior work you should verify for each non-T5 idea.

Frequently Asked Questions about idea-brainstorm

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

FAQPage Schema
How do I generate novel research ideas for a computer architecture problem?

Generate novel research ideas by applying divergence lenses like constraint relaxation and cross-domain analogy to a topic, then verifying each idea against prior work to assign qualitative novelty tiers from T1 to T5.

What is the best way to verify research novelty before starting a literature review?

Verify research novelty by contrasting candidate ideas against closest prior work and identifying what must be checked, defaulting unknown novelty downward to ensure only genuinely original directions are tiered highly.

Can I use brainstorming lenses for AI/ML and Nature-family science domains?

Apply brainstorming lenses such as strategic ignorance, scale shifts, primitive invention, and bisociation to expand research ideation across AI/ML and Nature-family science domains.

How do I rank research ideas by originality to spot the most promising directions?

Rank research ideas by originality by sorting them into qualitative novelty tiers from T1 (highest) to T5, allowing you to quickly spot the most original directions and filter out incremental ones.

Does idea generation include a way to test if a research direction is viable?

Idea generation includes an optional fail-fast handoff layer that adds an Idea Card with the cheapest decisive test to kill or confirm top research ideas before you commit resources.

What limitations exist when assigning novelty tiers to speculative tech ideas?

Novelty tier assignment defaults unknown novelty downward and requires novelty verification against closest prior work, meaning speculative tech ideas without verifiable contrast points will likely receive lower tiers.