| name | paper-fetch |
| description | Fetch paper metadata from lightweight academic sources like arXiv and save normalized paper JSON records. |
| metadata | {"openclaw":{"emoji":"📥","requires":{"bins":["python3"]}}} |
paper-fetch
Use this skill when the user wants to search papers by topic, keyword, or free-text query.
Workflow
- Run
python3 /root/openclaw-dev/workspace/scripts/fetch_papers.py "<query>" --max-results <N>
- Read the JSON output
- Use returned
saved_to paths as inputs for later reflection
Notes
- This MVP uses arXiv first for stability and low dependency overhead.
- Outputs are saved under
/root/openclaw-dev/workspace/data/papers/.