fetch-research-papers

name: fetch-research-papers description: > Fetch and retrieve the latest research papers from Scopus. Use this skill when asked to: find recent papers, search for research, get papers on a topic, fetch paper details by DOI, summarize new research, discover what's new in a field, or save papers to Obsidian.

Fetch Research Papers Skill

This skill uses the Scopus API (Elsevier) to search for and retrieve research papers.

Prerequisites

• Environment variable SCOPUS_API_KEY must be set.
• Python 3 must be available.
• The helper script is at: ~/.copilot/skills/fetch-research-papers/fetch_papers.py

If SCOPUS_API_KEY is not set, remind the user to run:

export SCOPUS_API_KEY="your-key-here"

Get a free API key at: https://dev.elsevier.com/

Searching for Papers

Use the script to search for papers:

python3 ~/.copilot/skills/fetch-research-papers/fetch_papers.py \
  --query "<topic or keywords>" \
  --count 10 \
  --sort newest

Sort options:

• newest — most recently published (default)
• cited — most cited
• relevance — best keyword match
• oldest — oldest first

Filter by year:

python3 ~/.copilot/skills/fetch-research-papers/fetch_papers.py \
  --query "<topic>" --count 10 --sort newest --year-from 2023

Advanced Scopus query syntax (more precise):

# Search in title, abstract, keywords
python3 ... --query "TITLE-ABS-KEY(transformer architecture)"

# Specific journal
python3 ... --query "SOURCE(Nature) AND TITLE-ABS-KEY(climate change)"

# Specific author
python3 ... --query "AUTH(Hinton, Geoffrey)"

# Publication year range
python3 ... --query "TITLE-ABS-KEY(BERT) AND PUBYEAR > 2022"

Fetch a Single Paper by DOI

python3 ~/.copilot/skills/fetch-research-papers/fetch_papers.py \
  --doi "10.1016/j.neunet.2022.01.001"

Saving a Paper to Obsidian

To fetch a paper and get it formatted as an Obsidian note:

python3 ~/.copilot/skills/fetch-research-papers/fetch_papers.py \
  --doi "<DOI>" --obsidian

Then save the output to the vault:

python3 ~/.copilot/skills/fetch-research-papers/fetch_papers.py \
  --doi "<DOI>" --obsidian > "/Users/yuxinli/Documents/Vault/Research/Papers/<Year> - <Short Title>.md"

You can also combine this with the obsidian-vault skill to organize and save notes intelligently.

Typical Workflow

When the user asks "find me the latest papers on X":

1. Run the fetch script with relevant keywords
2. Present the results clearly — title, authors, year, journal, DOI, abstract snippet
3. Ask if the user wants any of them saved to their Obsidian vault
4. If yes, use --obsidian flag and save to ~/Documents/Vault/Research/Papers/

When the user asks "find papers and save them to Obsidian":

1. Fetch papers using the script
2. For each selected paper, generate the Obsidian note with --obsidian
3. Save to /Users/yuxinli/Documents/Vault/Research/Papers/<Year> - <Short Title>.md
4. Optionally append a line to today's daily log in the vault

Rate Limits

The free Scopus API key allows:

• 20,000 requests/week for search
• 5,000 requests/week for abstract retrieval
• Results limited to 25 per query for non-institutional access

Keep --count at 10–25 for best results.