About this skill

Problem Background

Research ideation often stalls at specific points: an idea is still vague and hard to structure; a single paper does not make it clear whether it can grow into a publishable direction; assumptions, baselines, and ablations lack systematic stress-testing before experiments; and contributions become scattered during writing. This skill reframes those tasks into reusable prompt templates, so the AI can take on specific research collaboration roles instead of giving generic advice.

How It Works

It provides 18 prompts across 6 workflow stages:
- Idea exploration: deep discussion, Socratic probing, stress testing, and research gap discovery.
- Literature-driven ideation: extract new directions from one paper, synthesize multiple papers, or extrapolate trends.
- Critical review: simulate top-conference reviewers, challenge assumptions, and audit implicit premises.
- Experiment design: plan experiment blueprints, ablation studies, and baseline selection.
- Interdisciplinary fusion: explore AI for Science, cross-domain transfer, and physics-informed machine learning.
- Writing support: structure paper introductions and distill contribution points.

Each prompt file usually includes a System Prompt, User Prompt template, example dialogue, and usage tips. English and Chinese versions are organized by category root and zh/ subdirectories, making it easy to select by language.

Boundaries

This skill is best used for early-stage idea polishing, literature-based direction checks, and experiment planning with an AI collaborator. It is not an automatic literature review system, code executor, or paper-writing service. Users should still provide real research context, constraints, prior results, and paper details; otherwise the output can remain generic.

Screenshots

Use Cases

Using a vague AI research idea to run Socratic follow-up questions with AI and identify open research gaps.
After reading a target paper, extracting follow-up research directions from its methods, limitations, and assumptions.
Before running experiments, challenging assumptions with a simulated reviewer and designing ablations plus baseline comparisons.
When drafting an introduction, structuring background, research gap, and paper contributions into a clearer narrative.

Best For

Graduate students refining an AI-for-Science topic need to turn vague intuitions into discussable research questions.
Algorithm researchers preparing top-conference experiments need structured assumption audits, ablations, and baseline choices.
Research groups synthesizing multiple papers need prompts to turn cross-paper overlaps into new directions.
Researchers writing introductions need to organize the research gap, related-work boundary, and contribution statement.

Related Skills

Cold Shower Gate

@user_2da54b45/cold-shower

Switch AI from a polite executor into a skeptical thinking partner, using assumption checks, pre-mortems, and evidence gates to expose weak ideas early.

Professional 2026.08.29 ↓ 280