domain-paper-research

---
name: domain-paper-research
description: Conduct comprehensive domain-specific research paper discovery and compilation. Search across skills, Obsidian vault, and knowledge graph to find relevant papers, then compile into a structured collection with reading order, concept maps, and design decision matrices. Use when user asks for papers on a specific technical domain or wants to build a literature review for a project.
---

# Domain Paper Research

Conduct comprehensive research paper discovery and compilation for a specific technical domain.

## When to Use

- User asks for papers on a specific topic (e.g., "papers for building an AI employee platform")
- Need to compile a literature review for a project
- Want to understand the research landscape for a technical area
- Building a reading list for a new domain

## Workflow

### Step 1: Search Multiple Sources

Search across all available sources simultaneously:

**A. Skills Search**
```bash
# List skills in relevant categories
skills_list --category research
skills_list --category openclaw-imports

# View relevant skills
skill_view <skill_name>
```

**B. Obsidian Vault Search**
```bash
# Search by filename
find "$VAULT" -name "*.md" -iname "*<keyword>*"

# Search by content
grep -rli "<keyword>" "$VAULT" --include="*.md"
```

**C. Knowledge Graph Search** (if available)
```python
kg_search_papers("<query>", limit=20)
kg_recommend_skills("<topic>", min_utility=0.8)
```

### Step 2: Categorize Findings

Group papers into logical categories:

| Category | Description | Example |
|----------|-------------|---------|
| Architecture | System-level design | Orchestration, Infrastructure |
| Coordination | Multi-agent patterns | Organization, Communication |
| Memory | State management | Long-term, Short-term, Multimodal |
| Tools | External capabilities | Reliability, Schema, Integration |
| Roles | Agent specialization | Assignment, Expertise |
| Engineering | Development methodology | Best practices, Bootstrapping |

### Step 3: Extract Key Information

For each paper, capture:
- **Title & Authors**
- **Source** (arXiv ID, conference, report ID)
- **Core Insight** (1-2 sentences)
- **Key Contributions** (bullet points)
- **Performance Metrics** (if applicable)
- **Applicable Scenarios**
- **Tags**

### Step 4: Create Structured Output

Generate a markdown document with:

1. **Overview** - Domain summary
2. **Categorized Paper List** - Grouped by topic
3. **Reading Order** - Phased approach (Week 1-2, etc.)
4. **Concept Map** - Visual relationships
5. **Design Decision Matrix** - Compare alternatives
6. **Open Questions** - Future exploration

### Step 5: Cross-Link in Obsidian

Update relevant area/project notes:
```markdown
### <Domain>
- [[Paper Collection Note]]

## 📚 Resources
- [[Paper Collection Note|<Description>]]
```

## Output Template

```markdown
# <Domain> - Paper Collection

## 📊 Overview
[Domain description and scope]

---

## [Category 1]

### [[Paper 1]] ⭐⭐⭐⭐⭐
- **Source**: [arXiv ID / Conference / Report]
- **Authors**: [Names]
- **Core Insight**: [Key finding]
- **Key Contributions**:
  - [Contribution 1]
  - [Contribution 2]
- **Performance**: [Metrics if applicable]
- **Applicable Scenarios**: [When to use]
- **Tags**: #[tag1] #[tag2]

---

## 📚 Recommended Reading Order

### Phase 1: Foundation (Week 1-2)
1. [Paper A] - [Why first]
2. [Paper B] - [Prerequisites]

### Phase 2: [Topic] (Week 3-4)
...

## 🔗 Concept Map

```
[Visual diagram of relationships]
```

## 🎯 Design Decision Matrix

| Problem | Option A | Option B | Recommendation |
|---------|----------|----------|----------------|
| [Question] | [Approach 1] | [Approach 2] | [Rationale] |

## 📝 Open Questions

1. [Question for future exploration]
2. ...
```

## Example Domains

- **AI Employee Platform**: Orchestration, multi-agent, memory, tools, roles
- **RAG Systems**: Retrieval, embedding, chunking, reranking, evaluation
- **Agent Memory**: Short-term, long-term, multimodal, forgetting
- **Multi-Agent Coordination**: Hierarchy, voting, consensus, communication

## Tips

1. **Prioritize high-utility papers** (utility >= 0.85 if using knowledge graph)
2. **Include both academic and industry sources** (arXiv + Gartner/etc)
3. **Create bidirectional links** in Obsidian for discoverability
4. **Add reading phases** to make the collection actionable
5. **Include design matrices** to help with architectural decisions

## References

This skill was developed from compiling papers for:
- AI employee platforms
- Agent engineering harness
- Multi-agent systems
- Agent memory and orchestration

## Activation Keywords

- "domain-paper-research"
- "domain paper research"
- "use domain paper research"
- "domain paper research help"
- "domain paper research tool"

## Tools Used

- `Read` - Read existing files and documentation
- `Write` - Create new files and documentation
- `Bash` - Execute commands when needed

## Instructions for Agents

1. Identify user's intent and specific requirements
2. Gather necessary context from files or user input
3. Execute appropriate actions using available tools
4. Provide clear results and suggest next steps

## Examples

### Basic Domain Paper Research usage
```
User: "Help me with domain paper research"
→ Understand requirements → Execute actions → Provide results
```

### Advanced usage
```
User: "I need detailed domain paper research assistance"
→ Clarify scope → Provide comprehensive solution → Follow up
```