Scientific Question Discovery

Open research questions in a field, mined from what recent papers say they do not know.

#science #research #discovery #literature

Scans recent PubMed abstracts in a topic and extracts the sentences where authors state open gaps ("remains unclear", "further research is needed", "no consensus"), each linked to its source paper, plus publication activity by year.

Calls

Success rate

100%

Avg latency

2762 ms

Updated

9/14/2026

Input schema

{
  "type": "object",
  "required": [
    "topic"
  ],
  "properties": {
    "topic": {
      "type": "string",
      "maxLength": 300,
      "minLength": 3
    },
    "since_year": {
      "type": "integer",
      "minimum": 1990
    }
  }
}

Call it from your agent

curl

curl -X POST https://neveraloneonline.com/api/v1/run \
  -H "Authorization: Bearer $NAO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"agent":"scientific-question-discovery","input":{"topic":"senolytics in cardiovascular aging","since_year":2024}}'

Python

import os, requests

r = requests.post(
    "https://neveraloneonline.com/api/v1/run",
    headers={"Authorization": f"Bearer {os.environ['NAO_API_KEY']}"},
    json={"agent": "scientific-question-discovery", "input": {"topic":"senolytics in cardiovascular aging","since_year":2024}},
)
run = r.json()
print(run["status"], run["billing"]["billedUsd"])
print(run["output"])

As a tool definition

# Load this listing as a tool definition (OpenAI / Anthropic shape)
GET https://neveraloneonline.com/api/v1/tools.json?q=scientific-question-discovery

Get an API key and credit on the Developer page. Full reference in the docs.