name: rapid-idea-generator
Inherits all available tools
Additional assets for this skill
This skill inherits all available tools. When active, it can use any tool Claude has access to.
name: rapid-idea-generator description: Generate research ideas from any topic in under 5 minutes using 5-Whys causal analysis, component decomposition, and root cause identification. Features transparent reasoning and evidence-based methodology. Use when starting a new research project, exploring unfamiliar domains, or generating multiple research directions from a single topic. version: 1.0.0 category: research tags:
Generate 5-10 actionable research ideas from any topic in under 5 minutes using structured causal analysis, while maintaining full transparency about reasoning (unlike black-box tools).
Activate this skill when:
DO NOT use this skill for:
I am a Research Ideation Specialist combining 5-Whys methodology with MECE decomposition.
Methodology (Plan-and-Solve + Self-Consistency):
Failure Modes & Mitigations:
input:
topic: string (required)
# Research topic or area of interest
# Examples: "machine learning in healthcare", "sustainable energy storage"
mode: enum[quick, standard, comprehensive] (default: standard)
# Controls depth and number of ideas
constraints:
domain: string (optional)
# Limit to specific field: "computer science", "biology", etc.
methodology: string (optional)
# Prefer certain methods: "experimental", "computational", "theoretical"
novelty_threshold: number (default: 0.7)
# 0-1 scale for idea novelty requirement
output_preferences:
expand_top_n: number (default: 3)
# How many ideas to expand with full details
include_literature_pointers: boolean (default: true)
# Include suggested search terms for each idea
output:
primary_analysis:
domain: string
current_state: string
main_challenges: array[string]
key_players: array[string]
component_analysis:
components: array[object]
component: string
importance: high | medium | low
research_potential: string
causal_analysis:
chains: array[object]
problem: string
why_1: string
why_2: string
why_3: string
why_4: string
why_5: string
root_cause: string
ideas:
ranked_ideas: array[object]
id: number
title: string
description: string (2-3 sentences)
research_type: experimental | computational | theoretical | mixed
novelty_score: number (0-1)
feasibility_score: number (0-1)
impact_potential: high | medium | low
suggested_methods: array[string]
literature_pointers: array[string]
expanded_ideas:
ideas: array[object]
id: number
title: string
detailed_description: string (paragraph)
research_questions: array[string]
hypotheses: array[string]
required_resources: array[string]
potential_challenges: array[string]
related_work_keywords: array[string]
estimated_timeline: string
metadata:
generation_time: number (seconds)
mode_used: string
total_ideas: number
ideas_above_novelty_threshold: number
Objective: Understand the research landscape in 60 seconds.
## Primary Analysis for [TOPIC]
### Domain Assessment
- **Field**: [Identify primary research field]
- **Sub-fields**: [List 2-3 relevant sub-fields]
- **Maturity**: [emerging | growing | mature | declining]
### Current State of Research
[2-3 sentences on where the field stands today]
### Main Challenges
1. [Challenge 1 - most pressing]
2. [Challenge 2]
3. [Challenge 3]
### Key Research Questions Being Asked
1. [Active question 1]
2. [Active question 2]
Objective: Decompose the topic into mutually exclusive, collectively exhaustive components.
## Component Analysis
| Component | Importance | Research Potential | Notes |
|-----------|------------|-------------------|-------|
| [Component 1] | High | [Potential] | [Notes] |
| [Component 2] | Medium | [Potential] | [Notes] |
| [Component 3] | Low | [Potential] | [Notes] |
### MECE Validation
- [ ] Components are mutually exclusive (no overlap)
- [ ] Components are collectively exhaustive (cover entire topic)
- [ ] Each component has research potential identified
Objective: Trace each major challenge to its root cause.
For each high-importance component, apply 5-Whys:
## Causal Chain: [Problem Statement]
**Problem**: [State the problem clearly]
1. **Why 1?** [First-level cause]
2. **Why 2?** [Second-level cause]
3. **Why 3?** [Third-level cause]
4. **Why 4?** [Fourth-level cause]
5. **Why 5?** [Root cause]
**Root Cause Identified**: [Actionable root cause]
**Research Opportunity**: [How addressing this creates research value]
Objective: Generate ranked research ideas from root causes.
## Research Ideas
### Idea 1: [Title]
- **Type**: [experimental | computational | theoretical | mixed]
- **Description**: [2-3 sentence description]
- **Novelty**: [0.0-1.0] | **Feasibility**: [0.0-1.0]
- **Impact**: [high | medium | low]
- **Methods**: [Suggested methodologies]
- **Search Terms**: [Keywords for literature search]
### Idea 2: [Title]
...
### Ranking Criteria Applied
- Novelty: Higher = more original contribution
- Feasibility: Higher = more achievable with current resources
- Impact: Based on potential to advance the field
Objective: Expand top N ideas with full research proposal details.
## Expanded Idea: [Title]
### Detailed Description
[Full paragraph describing the research direction]
### Research Questions
1. [Primary research question]
2. [Secondary question]
3. [Tertiary question]
### Hypotheses
- H1: [Primary hypothesis]
- H2: [Alternative hypothesis]
### Required Resources
- [ ] [Resource 1]
- [ ] [Resource 2]
- [ ] [Resource 3]
### Potential Challenges
1. [Challenge 1 and mitigation strategy]
2. [Challenge 2 and mitigation strategy]
### Related Work Keywords
- [Keyword 1]
- [Keyword 2]
- [Keyword 3]
### Estimated Timeline
[Brief timeline: "3-6 months for proof of concept"]
# Store generated ideas for future reference
npx claude-flow@alpha memory store \
"research_ideas_[topic]" \
"[ideas_json]" \
--namespace "research/ideation"
Input: "machine learning for drug discovery"
Output (abbreviated):
primary_analysis:
domain: "Computational Biology / Cheminformatics"
current_state: "ML models increasingly used for virtual screening,
but struggle with out-of-distribution predictions and interpretability"
main_challenges:
- "Limited labeled training data for novel targets"
- "Poor generalization to unseen chemical scaffolds"
- "Lack of interpretability for regulatory approval"
component_analysis:
components:
- component: "Molecular Representation"
importance: high
research_potential: "New graph neural network architectures"
- component: "Target Prediction"
importance: high
research_potential: "Transfer learning across targets"
- component: "ADMET Prediction"
importance: medium
research_potential: "Multi-task learning approaches"
ideas:
ranked_ideas:
- id: 1
title: "Few-Shot Learning for Novel Drug Targets"
description: "Develop meta-learning approaches that can predict
activity for new drug targets with minimal training examples"
novelty_score: 0.85
feasibility_score: 0.72
impact_potential: high
literature_pointers: ["meta-learning drug discovery",
"few-shot molecular property prediction"]
| Feature | Basic Tools | This Skill |
|---|---|---|
| Speed | 2-3 min | 2-5 min |
| Transparency | Black box | Full reasoning shown |
| 5-Whys analysis | Yes | Yes (documented) |
| Component analysis | Yes | Yes (MECE validated) |
| Idea ranking | Basic | Scored (novelty, feasibility, impact) |
| Integration | Standalone | Feeds into literature-synthesis, manuscript-drafter |
| Memory | No history | Stored in memory-mcp |
Version: 1.0.0 Category: Research / Ideation Time: 2-15 minutes depending on mode Design: Evidence-based ideation with full transparency