OpenScience
VerifiedRuns a research agent for scientists that reviews literature, writes and runs analysis code, and reproduces experiments, from a desktop app, browser workspace, or CLI.
- Access
- CLIAPI
- Auth
- API Key
- Skills
- 5
Skills
Literature Review
Searches scientific sources, compares findings, and saves cited evidence, with connectors to databases such as PubMed, arXiv, and UniProt.
Data Analysis
Inspects input data, writes and runs analysis code in shell, Python, or R kernels, and produces figures and reports in the project folder.
Experiment Reproduction
Agrees on a claim, its prerequisites, and a budget with you, then reruns the experiment and compares the measured results to the claim.
Remote Compute
Sends long or GPU-bound jobs to Modal, SSH hosts, or Slurm and PBS clusters, with each dispatch approved by you before it runs.
Headless Runs
Runs a single research turn from a script or pipeline with openscience run, and continues the same session with the --continue flag.
Related Agents
STORM
by Stanford OVAL
Researches a topic and writes a cited, Wikipedia-length report by simulating multi-perspective question asking.…
ToolUniverse
by Harvard MIMS Lab
Curates a large collection of scientific tools and MCP servers for AI scientists, exposing biomedical and research APIs…
LinqAlpha
by LinqAlpha
Run research agents over filings, transcripts, and news for buy-side investing — encodes an analyst research framework…
OpenAlex AI Agent Connector
by OpenAlex
Lets AI assistants search and analyze OpenAlex scholarly works, authors, institutions, and citations over a remote MCP…