Get new research papers on my topic each week
New papers and preprints on your topic, the most relevant first, with a plain summary of each.
- On a schedule: Every Friday at 8:00
What it does
The task
Search arXiv, PubMed and Semantic Scholar for papers and preprints from the last 7 days about topic. Also include new papers that cite these papers or come from these labs: seeds. Remove duplicates of the same work. Send me the 10 most relevant, each with the title, the institutions, the venue, a 2-line summary of the method and the main result, and the link. Mark the preprints that are not peer reviewed.
- It works in arXiv, PubMed and Semantic Scholar, in the workspace you pick, and signs in with passwords from your Cuiar vault.
- As a routine it runs every Friday at 8:00 (cron
0 8 * * 5).
What you give it
| Blank | What to give | Example |
|---|---|---|
| topic | Your research topic in a few words | Retrieval for long-context language models |
| seeds | Key papers or labs to follow | the paper 'Memory tokens for retrieval' and the Westvale AI Lab |
What you get back
An example run with the example values above. Your run uses your own data.
New papers on retrieval for long-context language models: 10 of 46 found
| 1. 'Sparse memory tokens for million-token retrieval' | Westvale AI Lab | arXiv preprint, not peer reviewed |
|---|
It adds 64 memory tokens to each chunk of text. Recall on a 1-million-token test goes up by 12 points. Link
| 2. 'When retrieval hurts: a study of 9 long-context models' | Northgate University | accepted at a 2027 conference |
|---|
Retrieval helps under 128,000 tokens and gives worse answers above it for 6 of 9 models. Link
| 3. 'Chunk size and answer quality in medical question answering' | Harbor Medical School | PubMed, peer reviewed |
|---|
Chunks of 512 tokens gave the best answers on 3 medical tests. Link
7 more papers in the e-mail
Run it in Cuiar
- 01
Open it
Click Use in Cuiar on this page. The side panel opens with the task in the composer.
- 02
Fill in 2 blanks
Type topic and seeds. Tab moves to the next blank.
- 03
Run it, then repeat it
Run sends it once. To run it every Friday at 8:00, make a routine on cuiar.com/routines or with the CLI below.
As a routine, from the terminal
cuiar routines create "Get new research papers on my topic each week" --schedule "0 8 * * 5" --time-zone Europe/Berlin \
"Search arXiv, PubMed and Semantic Scholar for papers and preprints from the last 7 days about Retrieval for long-context language models. Also include new papers that cite these papers or come from these labs: the paper 'Memory tokens for retrieval' and the Westvale AI Lab. Remove duplicates of the same work. Send me the 10 most relevant, each with the title, the institutions, the venue, a 2-line summary of the method and the main result, and the link. Mark the preprints that are not peer reviewed."Run it from your AI agent
Claude Code, Codex, Cursor and other MCP clients reach Cuiar at https://mcp.cuiar.com/mcp. Add the server one time, then ask for the task in plain words, or call create_routine yourself.
1. Add Cuiar to Claude Code
claude mcp add --transport http cuiar https://mcp.cuiar.com/mcp2. Ask your agent
Use the Cuiar tool create_routine to run my "Get new research papers on my topic each week" template with topic = Retrieval for long-context language models; seeds = the paper 'Memory tokens for retrieval' and the Westvale AI Lab.Setup for other agents: AI agents.
The tool call
{
"name": "create_routine",
"arguments": {
"name": "Get new research papers on my topic each week",
"task": "Search arXiv, PubMed and Semantic Scholar for papers and preprints from the last 7 days about Retrieval for long-context language models. Also include new papers that cite these papers or come from these labs: the paper 'Memory tokens for retrieval' and the Westvale AI Lab. Remove duplicates of the same work. Send me the 10 most relevant, each with the title, the institutions, the venue, a 2-line summary of the method and the main result, and the link. Mark the preprints that are not peer reviewed.",
"schedule": "0 8 * * 5",
"time_zone": "Europe/Berlin"
}
}Related templates
All monitoring templatesResearchOdyssey
Write a literature review on a research question
Recent studies grouped by theme, what they agree on, where the evidence is weak, and a full reference list.
You start it: When you start it- Google Scholar
- Semantic Scholar
- PubMed
- +1 more: Google Docs
Research
Explain a research paper in plain words
The question, the method, the results and the limits of one paper, for a reader with no background.
You start it: When you start it- arXiv
- Semantic Scholar
ResearchOdyssey
Find what studies say about a treatment
The best evidence for and against, the size of the effect, the side effects and questions for your doctor.
You start it: When you start it- PubMed
- Google Scholar
Research
Find clinical trials for a condition
Trials that recruit now in your area, with the main criteria, the sites and the link to each.
You start it: When you start it- Any government site
- PubMed
Data
Extract study results into a table
Design, sample size, outcome and effect size with confidence interval from each paper, in one sheet.
You start it: When you start it- PubMed
- Google Scholar
- Google Sheets
Monitoring
Tell me when a price drops
A message only when the price goes below your target.
On a schedule: Every day at 9:00- Any online store