/answering-research-questions

name: Answering Research Questions
description: Main orchestration workflow for systematic literature research - search, evaluate, traverse, synthesize
when_to_use: When user asks a research question. When user wants to find specific data in literature. When starting comprehensive literature review. When user says "find papers about" or "what is known about".
version: 1.0.0

<!-- ╔══════════════════════════════════════════════════════════════╗ ║ 本文件为开源 Skill 原始文档，收录仅供学习与研究参考 ║ ║ CoPaper.AI 收集整理 | https://copaper.ai ║ ╚══════════════════════════════════════════════════════════════╝

来源仓库: https://github.com/kthorn/research-superpower 项目名称: research-superpower 开源协议: MIT License 收录日期: 2026-04-02

声明: 本文件版权归原作者所有。此处收录旨在为社会科学实证研究者 提供 AI Agent Skills 的集中参考。如有侵权，请联系删除。 -->

Answering Research Questions

Overview

Orchestrate the complete research workflow from query to findings.

**Core principle:** Systematic, trackable, comprehensive. Search → Evaluate → Traverse → Synthesize.

**Announce at start:** "I'm using the Answering Research Questions skill to find [specific data] about [topic]."

The Process

Phase 1: Parse Query

Extract from user's request:

**Keywords:**

• Main concepts (e.g., "BTK inhibitor", "selectivity")
• Synonyms and alternatives (e.g., "Bruton tyrosine kinase")
• Related terms (e.g., "off-target", "kinase panel")

**Data types needed:**

• Specific measurements (IC50, KD, EC50, etc.)
• Methods or protocols
• Structures or sequences
• Results or conclusions

**Constraints:**

• Date ranges
• Specific compounds/targets
• Organisms or systems
• Publication types

**Ask clarifying questions if needed:**

• "Are you looking for in vitro or in vivo data?"
• "Any specific time frame?"
• "Which kinases are you most interested in?"
• **"What email address should I use for Unpaywall API requests?"** (Required for finding open access papers)

Phase 2: Initialize Research Session

**Propose folder name:**

research-sessions/YYYY-MM-DD-brief-description/

Example: `research-sessions/2025-10-11-btk-inhibitor-selectivity/`

**Show proposal to user:**

📁 Creating research folder: research-sessions/2025-10-11-btk-inhibitor-selectivity/
   Proceed? (y/n)

**Create folder structure:**

mkdir -p "research-sessions/YYYY-MM-DD-description"/{papers,citations}

**Initialize files:**

**Core files (always create these):**

**papers-reviewed.json:**

{}

**citations/citation-graph.json:**

{}

**SUMMARY.md:**

# Research Query: [User's question]

**Started:** YYYY-MM-DD HH:MM
**Keywords:** keyword1, keyword2, keyword3
**Data types sought:** IC50 values, selectivity data, synthesis methods

---

## Highly Relevant Papers (Score ≥ 8)

Papers scored using `evaluating-paper-relevance` skill:
- Score 0-10 based on: Keywords (0-3) + Data type (0-4) + Specificity (0-3)
- Score ≥ 8: Highly relevant with significant data
- Score 7: Relevant with useful data
- Score 5-6: Possibly relevant
- Score < 5: Not relevant

(Papers will be added here as found)

Example format:
### [Paper Title](https://doi.org/10.1234/example)
**DOI:** [10.1234/example](https://doi.org/10.1234/example) | **PMID:** [12345678](https://pubmed.ncbi.nlm.nih.gov/12345678/)

---

## Relevant Papers (Score 7)

(Papers will be added here as found)

---

## Possibly Relevant Papers (Score 5-6)

(Noted for potential follow-up)

---

## Search Progress

- Initial PubMed search: X results
- Papers reviewed: Y
- Papers with relevant data: Z
- Citations followed: N

---

## Key Findings

(Synthesized findings will be added as research progresses)

**CRITICAL: Always use clickable markdown links for DOIs and PMIDs**

**Auxiliary files (for large searches >100 papers):**

See `evaluating-paper-relevance` skill for guidance on when to create:

• **README.md** - Project overview, methodology, file inventory
• **TOP_PRIORITY_PAPERS.md** - Curated priority list organized by tier
• **evaluated-papers.json** - Rich structured data for programmatic access

For small searches (<50 papers), stick to core files only. For large searches (>100 papers), auxiliary files add significant organizational value.

Phase 3: Search Literature

**Use searching-literature skill:**

1. Construct PubMed query from keywords 2. Execute search (start with 100 results) 3. Save results to `initial-search-results.json` 4. Report: "🔎 Found N papers matching query"

Phase 4: Evaluate Papers

**Use evaluating-paper-relevance skill:**

For each paper: 1. Check papers-reviewed.json (skip if already processed) 2. Stage 1: Score abstract (0-10) 3. If score ≥ 7: Stage 2 deep dive 4. Extract findings to SUMMARY.md 5. Download PDF and supplementary if available 6. **Update papers-reviewed.json (for ALL papers, even low-scoring ones)** 7. If score ≥ 7: proceed to Phase 5 for this paper

**CRITICAL: Add every paper to papers-reviewed.json regardless of score. This prevents re-review and tracks complete search history.**

**Report progress for EVERY paper:**

📄 [15/100] Screening: "Paper Title"
   Abstract score: 8 → Fetching full text...
   ✓ Found IC50 data for 8 compounds
   → Added to SUMMARY.md

📄 [16/100] Screening: "Another Paper"
   Abstract score: 3 → Skipping (not relevant)

📄 [17/100] Screening: "Third Paper"
   Abstract score: 7 → Relevant, adding to queue...

**Every 10 papers, give summary update**

Phase 5: Traverse Citations

**Use traversing-citations skill:**

For papers scoring ≥ 7: 1. Get references (backward) 2. Get citations (forward) 3. Filter for relevance (score ≥ 5) 4. Add to processing queue 5. Evaluate queued papers (return to Phase 4)

**Report progress:**

🔗 Following citations from highly relevant paper
   → Found 12 relevant references
   → Found 8 relevant citing papers
   → Adding 20 papers to queue

Phase 6: Checkpoint

**Check after:**

• Every 50 papers reviewed
• Every 5 minutes of processing
• Qu