Actions
Jump to:
searchdownload
Search arXiv for papers matching a query and return title, authors, id, PDF link and summary for each match.
Inputs
| Name | Type | Required / default | Values and notes |
query |
string |
required
|
|
max_results |
int |
optional, default 10
|
|
Outputs
| Name | Type | Notes |
papers |
json |
|
count |
int |
|
Download a paper
download
Download one paper's PDF into a folder, given its arXiv id or an arxiv.org URL (abs or pdf link).
Inputs
| Name | Type | Required / default | Values and notes |
id_or_url |
string |
required
|
|
out_dir |
path |
required
|
|
filename |
string |
optional, default ""
|
|
Outputs
| Name | Type | Notes |
file |
path |
|
skipped |
bool |
|
bytes |
int |
|
Use it
Search arXiv
Command line
shell
python app.py --action search --params '{"query": "<query>"}'
Run from the app’s own folder. Single quotes work in bash and PowerShell; in cmd.exe use double quotes and escape the inner ones.
MCP tool
mcp
run_action("arxiv-fetcher", "search", {"query": "<query>"})
Workflow node
json
{
"id": "n0",
"app": "arxiv-fetcher",
"action": "search",
"params": {
"query": "<query>"
},
"inputs": {}
}
Download a paper
Command line
shell
python app.py --action download --params '{"id_or_url": "<id_or_url>", "out_dir": "D:/path/to/out_dir"}'
Run from the app’s own folder. Single quotes work in bash and PowerShell; in cmd.exe use double quotes and escape the inner ones.
MCP tool
mcp
run_action("arxiv-fetcher", "download", {"id_or_url": "<id_or_url>", "out_dir": "D:/path/to/out_dir"})
Workflow node
json
{
"id": "n0",
"app": "arxiv-fetcher",
"action": "download",
"params": {
"id_or_url": "<id_or_url>",
"out_dir": "D:/path/to/out_dir"
},
"inputs": {}
}
Example prompts
Things you can say to Claude, Cursor or Codex once the MCP server is registered.
- search arXiv for 'diffusion models' and show me the top 10
- find recent papers on transformer interpretability
- download arXiv paper 2301.00001 into D:/papers
- grab the PDF for https://arxiv.org/abs/1706.03762