---
title: "Usage Examples"
description: "Multi-step research workflows showing how ScholarXIV MCP tools compose together."
canonical_url: "https://scholarxiv.com/developers/docs/mcp/examples"
markdown_url: "https://scholarxiv.com/developers/docs/mcp/examples.md"
---

# Usage Examples
URL: /developers/docs/mcp/examples
LLM index: /llms.txt
Description: Multi-step research workflows showing how ScholarXIV MCP tools compose together.
Related: mcp, mcp/tools, mcp/quickstart

# Usage Examples

Real-world workflows showing how MCP tools compose together.

## Workflow 1: Discover and Read Papers

Find papers on a topic, read details, and bookmark the best ones.

```
1. search_papers(query: "transformer attention mechanisms", limit: 5)
   → Returns 5 paper titles and abstracts

2. get_paper(paper_id: "2401.01234")
   → Returns full metadata, abstract, authors, categories

3. bookmark_paper(paper_id: "2401.01234")
   → Bookmarked

4. like_paper(paper_id: "2401.01234")
   → Liked
```

**What the AI agent does:** Searches, evaluates abstracts, reads full details, and saves interesting papers for later.

## Workflow 2: Build a Collection

Create a curated collection and populate it with papers.

```
1. create_collection(
     title: "Attention Mechanisms",
     description: "Papers on attention and self-attention in neural networks",
     tags: ["attention", "transformers", "deep learning"]
   )
   → Collection created with ID

2. search_papers(query: "self-attention neural networks", limit: 10)
   → Returns 10 papers

3. add_paper_to_collection(collection_id: "...", paper_id: "2401.01234")
   → Paper added

4. add_paper_to_collection(collection_id: "...", paper_id: "2401.05678")
   → Paper added

5. update_collection(collection_id: "...", visibility: "public")
   → Collection is now public

6. generate_share_token(collection_id: "...", role: "viewer")
   → Share link generated
```

**What the AI agent does:** Creates a themed collection, populates it with relevant papers, makes it public, and generates a share link.

## Workflow 3: Research with AI

Use the AI research assistant for deep analysis.

```
1. research_chat(
     question: "What are the main challenges in applying transformers to protein folding?",
     is_deep_research: true
   )
   → AI searches papers, analyzes findings, provides synthesis

2. get_chat_history(chat_id: "...")
   → View the full conversation

3. search_papers(query: "protein structure prediction transformer")
   → Find specific papers mentioned

4. bookmark_paper(paper_id: "...")
   → Save key papers
```

**What the AI agent does:** Uses the AI research assistant for deep analysis, then saves the most relevant papers.

## Workflow 4: Collaborate on Collections

Share a collection with collaborators.

```
1. create_collection(title: "Team Reading List")
   → Collection created

2. add_paper_to_collection(collection_id: "...", paper_id: "...")
   → Papers added

3. generate_share_token(collection_id: "...", role: "editor")
   → Editor link generated (collaborators can add papers)

4. get_collection_members(collection_id: "...")
   → Check who has access
```

**What the AI agent does:** Creates a shared workspace, adds papers, and generates an editor link for collaborators.

## Workflow 5: Paper Discussion

Read comments on a paper and join the discussion.

```
1. get_paper(paper_id: "2502.03725")
   → Read the paper details

2. get_paper_comments(paper_id: "2502.03725")
   → See existing discussion

3. comment_on_paper(
     paper_id: "2502.03725",
     comment: "The approach to fluid control is innovative. Has anyone tested this on larger state spaces?"
   )
   → Comment posted

4. get_paper_comments(paper_id: "2502.03725")
   → See updated discussion
```

**What the AI agent does:** Reads the paper, checks community discussion, and contributes a thoughtful comment.

## Workflow 6: Explore What Others Are Reading

Discover popular collections and join interesting ones.

```
1. discover_collections(page: 1)
   → Browse random public collections

2. search_collections(query: "reinforcement learning")
   → Find collections on a specific topic

3. get_collection_detail(collection_id: "...")
   → See what papers are in a collection

4. join_collection(collection_id: "...")
   → Join as a viewer

5. get_recommended_collections(page: 1)
   → Get personalized recommendations
```

**What the AI agent does:** Explores the community, finds interesting collections, and joins them for ongoing reading.

## Tips for Effective MCP Usage

- **Be specific** in search queries — "transformer attention mechanisms 2024" beats "AI papers"
- **Use collections** to organize research by topic or project
- **Bookmark incrementally** — bookmark papers as you find them, not just at the end
- **Use research_chat** for synthesis questions, not simple lookups
- **Try deep research** (`is_deep_research: true`) for complex multi-step analysis

## Sitemap

See the full [sitemap](/sitemap.md) for all pages.
Docs-scoped sitemap: [/docs/sitemap.md](/docs/sitemap.md).
Well-known sitemap: [/.well-known/sitemap.md](/.well-known/sitemap.md).