AgentSkillsCN

research-ideation

基于文献空白生成结构化的假设。梳理现有研究，识别尚未解答的问题，生成有文献支撑的假设，并运用评分标准评估其可行性（数据、时间表、资金、方法）。

SKILL.md
--- frontmatter
name: research-ideation
description: |
  Structured hypothesis generation from literature gaps. Maps existing research,
  identifies unanswered questions, generates citation-backed hypotheses, and
  assesses feasibility using a rubric (data, timeline, funding, methods).
author: BK
version: 1.0.0
date: 2026-02-20

Research Ideation

Generate research ideas grounded in existing literature. Identifies gaps, formulates testable hypotheses, and evaluates feasibility.

When to Use This Skill

Trigger when user:

  • •Says "research ideas", "what should I study next"
  • •Asks for "hypothesis generation" or "gap analysis"
  • •Wants to "brainstorm research questions"
  • •Says "what's missing in the literature on..."
  • •Needs ideas for a "dissertation topic" or "new project"

Prerequisites

  • •Zotero MCP: For searching user's existing library
  • •WebSearch: For discovering recent work beyond Zotero

Workflow

Phase 1: Scope the Domain

Use AskUserQuestion to clarify:

  1. •Field/topic: What broad area? (e.g., "social media and political polarization")
  2. •Existing knowledge: What do you already know? What have you published?
  3. •Constraints: Methods you can use, data you can access, timeline
  4. •Goal: Dissertation, journal article, grant proposal, pilot study?

Phase 2: Map the Literature

  1. •Search user's Zotero library for existing collections in this area
  2. •Search externally for recent work (last 2-3 years)
  3. •Identify:
    • •Established findings: What is well-documented and replicated?
    • •Active debates: Where do researchers disagree?
    • •Methodological gaps: What methods haven't been applied to this question?
    • •Population gaps: Who hasn't been studied?
    • •Temporal gaps: What time periods or contexts are missing?

Phase 3: Generate Hypotheses

For each identified gap, generate:

markdown
## Research Idea [N]: [Title]

**Research Question:** [Specific, testable question]

**Hypothesis:** [Directional prediction with theoretical justification]

**Grounded in:**
- [Citation 1]: Found X, but did not examine Y
- [Citation 2]: Called for future research on Z

**Proposed Method:** [Brief approach — survey, experiment, secondary data, etc.]

**Expected Contribution:** [What this would add to the field]

Generate 3-5 ideas, ranging from incremental to ambitious.

Phase 4: Feasibility Assessment

Score each idea on a rubric:

CriterionScore (1-5)Notes
Data availabilityExisting data? Need to collect?
Method feasibilityDo you have the skills/tools?
Timeline fitCan it be done in available time?
Funding fitMatches available/target funding?
Publication potentialWhich journals? How competitive?
NoveltyHow original relative to existing work?

Phase 5: Recommend Next Steps

For the top-rated idea(s):

  1. •Suggest a concrete first step (pilot study, data exploration, lit review)
  2. •Identify 2-3 key papers to read closely
  3. •Suggest potential co-authors or collaborators if relevant
  4. •Outline a rough timeline

Key Principles

  • •Every hypothesis must be backed by specific citations showing the gap
  • •Distinguish between "nobody has studied this" and "nobody has studied this well"
  • •Prioritize feasibility over novelty — a doable study beats a brilliant one that can't be executed
  • •Be honest about limitations (data access, method constraints)
  • •Generate a range: some safe/incremental ideas and some high-risk/high-reward ones