domains / science / research-ideation / scientific-brainstorming

scientific-brainstorming

Skill card

combined

The K-Dense copy is a formal, multi-participant facilitation protocol with strict operating rules, provenance tracking, and explicit hand-offs to other skills. The Vibe-Skills and OpenClaw copies are conversational, single-user ideation partners that use specific creative techniques (analogies, assumption reversal) and a different 5-phase workflow. They overlap in purpose but disagree on the interaction model (facilitated group vs. conversational partner) and the specific steps required, so crowning one would hide a real alternative.

Repositories

Other copies

Combined skill

# Scientific Brainstorming

## Purpose
Facilitate evidence-aware scientific ideation. This skill supports both formal multi-participant sessions and conversational single-user ideation. Treat all outputs as proposals, not findings.

## Operating Rules
1. **Label Claims**: Distinguish between ideas, assumptions, predictions, and located evidence.
2. **Independent Generation**: When facilitating a group, ensure participants generate ideas independently before sharing to avoid anchoring.
3. **Preserve Minority Views**: Record negative evidence, uncertainty, and dissenting opinions.
4. **Provenance**: Track the origin of ideas (human, AI-assisted, literature-inspired) without exposing confidential data.

## Workflow

### 1. Scope and Context
- **Formal**: Define the focal question, decision owner, in/out-of-scope topics, and constraints.
- **Conversational**: Ask open-ended questions to understand the researcher's current work, obstacles, and implicit assumptions.

### 2. Divergent Exploration
- **Formal**: Use a fixed time window for private, parallel idea generation.
- **Conversational**: Use creative techniques such as:
  - **Cross-Domain Analogies**: Draw parallels from other fields.
  - **Assumption Reversal**: Flip core assumptions.
  - **Scale Shifting**: Explore molecular to ecosystem scales.
  - **Constraint Removal**: Imagine unlimited resources or data.

### 3. Sharing and Clustering
- **Formal**: Use round-robin or pooled silent sharing. Cluster ideas by mechanism, population, or method.
- **Conversational**: Identify patterns and unexpected connections among the generated ideas.

### 4. Evaluation and Synthesis
- Define transparent criteria before scoring.
- Evaluate feasibility, novelty, and potential disconfirming evidence.
- Synthesize the most promising directions into concrete next steps (e.g., literature search, pilot experiments).

## Hand-offs
- For testable hypotheses, hand off to `hypothesis-generation`.
- For study architecture, use `experimental-design`.
- For ethics or regulatory review, route to authorized institutional reviewers.