idea-creator

Generate and rank publishable research ideas with experimental designs.

1|Updated May 14, 2026
One-click install
npx skills add https://github.com/lix965996-art/MMM --skill idea-creator-lix965996-art
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: idea-creator
Source: https://github.com/lix965996-art/MMM/tree/main/resources/app/skills/idea-creator
Command: npx skills add https://github.com/lix965996-art/MMM --skill idea-creator-lix965996-art

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It eliminates the struggle of converting a broad research direction into concrete, testable, and potentially publishable research ideas with clear experiments and evaluation plans.

Core Features & Use Cases

  • Research landscape + gap finding: Maps recent literature and organizes papers to identify structural gaps, assumptions, and open problems.
  • Ranked idea generation: Produces 8–12 ideas, filters them for feasibility/novelty/impact, then deep-validates the top candidates.
  • Pilot validation pipeline: Estimates GPU budget, runs short parallel pilots for the best 2–3 ideas when resources are available, and reranks based on empirical signal.

Use it when you want to explore a topic, brainstorm “what we can work on,” or find publishable directions for a research area.

Quick Start

Ask the AI to run idea-creator with your direction, for example: Generate research ideas for “federated learning with non-IID data” and include a pilot plan.

Frequently Asked Questions about idea-creator

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

FAQPage Schema
How do I generate publishable research ideas from a broad research direction?

To generate publishable research ideas, you provide a broad research direction as YAML input. The system surveys recent literature, identifies structural gaps, and produces 8–12 ranked, testable proposals tailored for publication.

What is the best way to check research idea novelty and feasibility before starting experiments?

Checking research idea novelty involves filtering generated hypotheses against existing literature surveys. The system deep-validates top candidates by evaluating their structural gaps, assumptions, and feasibility to ensure high-impact, novel proposals.

How do I plan pilot studies and estimate GPU budgeting for new research proposals?

Planning pilot studies requires estimating the GPU budget for the top 2–3 ranked ideas. The system facilitates this by preparing short parallel pilot validation pipelines and reranking the ideas based on the empirical signal gathered.

Can I use literature survey results to automatically brainstorm testable hypotheses?

Yes, you can use literature survey results to brainstorm testable hypotheses. The system maps recent papers to identify open problems and assumptions, automatically translating these structural gaps into concrete research proposals.

Does this research ideation workflow output a structured experiment planning document?

Yes, the research ideation workflow outputs a structured IDEA_REPORT.md document. This file contains the ranked ideas, experimental designs, and pilot validation plans necessary for moving forward with the research.