Research, fully automated. ClawsGO Science takes a question to discovery and writes it up as a cited paper. Its cloud lab has 500+ tools and 90+ datasets. Its long-horizon harness runs clusters of Claude and GPT agents for days. The next paradigm for science.
Hey Product Hunt 👋 I'm Annactswell, founder of ClawsGO Science.
We think the next paradigm for science is a lab that never closes: agents that take a question, run the experiments and write the paper while you sleep. To get there, an agent needs more than a model. It needs a lab.
Our team used Claude Code and Codex for research every day. They're great at code. But research isn't a repo. Ask one whether two gut bacteria cause depression, and it spends the first hour installing Mendelian randomisation packages and hunting for GWAS data. It runs on one vendor's models. And at the end, you get a diff, not a paper.
So we built the lab first. ClawsGO Science gives every task an always-on cloud workstation with terabytes of science software and data already installed. Claude and GPT agents work side by side for days, check their own work, then hand back a cited, compiled paper.
That gut bacteria question is in our gallery. ClawsGO came back with a 24-page paper: 17 figures, 10 tables, 59 references. See the full run: https://clawsgo.ai
🧪 A lab that's already set up
• 500+ tools and 90+ reference datasets across 12 fields, from GATK and GROMACS to Quantum ESPRESSO and QGIS
• An always-on workstation that keeps your files between tasks, with a full Linux desktop for GUI tools like Fiji
• Every search, command and file edit on record. Replay the whole run any time
🧭 Claude and GPT, working for days
• A harness built for long-horizon research keeps one task on track from question to paper
• Cluster mode: a lead agent splits the job and hands each part to the model that does it best
• Self-review: before you see the paper, it rereads the draft against the sources and the code, and fixes what it finds
📄 Ends with a cited paper
• A LaTeX workbench with source and PDF side by side, compiled in the cloud
• Figures you can mark up: circle what's wrong, ask for a fix, keep every version
📱 On every device: web, macOS, Windows, iOS and Android.
Who it's for: researchers whose work runs on code, tools and data, from genomes to galaxies. If you already use Claude Code or Codex for research, think of ClawsGO Science as the lab behind the agent.
Try it free at clawsgo.ai. You start with 1,500 credits, no card needed.
Three questions for you: what's the first research question you'd hand it? If you use Claude Code or Codex for research today, where does it get stuck? And which tool or dataset should be in the lab?
Thanks for checking us out. We're here all day and will answer every comment 🙏
running Claude and GPT agents side by side for days is a wild setup. what happens when they come back with different answers to the same question?
congrats on launch! I am wondering which area is the tool focus on reasearching?
Preinstalled tools are a big plus. Does a long job keep running if I close my laptop?
A compiled PDF from one research question sounds handy. Going to give the free credits a try.
The citation check is the feature I would test first. Can users see exactly how each reference was verified?
Package setup and LaTeX cleanup are usually where I lose time. Curious how much of that ClawsGO handles on its own.
The replay idea is what makes a multi-day run feel trustworthy to me — every search, command and file edit on record, rather than just a final paper to take on faith. If a replay shows a figure went wrong halfway through, can you branch from that step and rerun just the rest, or does fixing it mean starting the whole task over?
Congrats on the launch! Research is a great fit for an agent that covers the whole loop from finding papers to writing up results. How do you make sure the sources it cites are real and say what the report claims?
can i ask it to update a chart palette without changing the rest of the report?
A research agent that stays on one question for days across its own cloud lab is a completely different use of agents than the five minute loops I am used to, and citing sources makes the output checkable.
the self-review step is the part I keep thinking about - rereading a draft against sources and code will catch a citation mismatch, but what about when the agent just misread what the data actually shows in the first place? that feels like a harder failure mode than a wrong citation. does self-review catch that too, or is that still on the human to catch?
Can a private S3 bucket be mounted directly, or do files need to be uploaded?
About ClawsGO Science on Product Hunt
“The AI research agent with its own cloud lab”
ClawsGO Science launched on Product Hunt on November 1st, 2026 and earned 279 upvotes and 57 comments, earning #2 Product of the Day. Research, fully automated. ClawsGO Science takes a question to discovery and writes it up as a cited paper. Its cloud lab has 500+ tools and 90+ datasets. Its long-horizon harness runs clusters of Claude and GPT agents for days. The next paradigm for science.
ClawsGO Science was featured in Productivity (664.6k followers), Artificial Intelligence (482.7k followers) and Science (1.4k followers) on Product Hunt. Together, these topics include over 310.3k products, making this a competitive space to launch in.
Who hunted ClawsGO Science?
ClawsGO Science was hunted by Zac Zuo. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how ClawsGO Science stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Zac Zuo
Hey Product Hunt 👋 I'm Annactswell, founder of ClawsGO Science.
We think the next paradigm for science is a lab that never closes: agents that take a question, run the experiments and write the paper while you sleep. To get there, an agent needs more than a model. It needs a lab.
Our team used Claude Code and Codex for research every day. They're great at code. But research isn't a repo. Ask one whether two gut bacteria cause depression, and it spends the first hour installing Mendelian randomisation packages and hunting for GWAS data. It runs on one vendor's models. And at the end, you get a diff, not a paper.
So we built the lab first. ClawsGO Science gives every task an always-on cloud workstation with terabytes of science software and data already installed. Claude and GPT agents work side by side for days, check their own work, then hand back a cited, compiled paper.
That gut bacteria question is in our gallery. ClawsGO came back with a 24-page paper: 17 figures, 10 tables, 59 references. See the full run: https://clawsgo.ai
🧪 A lab that's already set up
• 500+ tools and 90+ reference datasets across 12 fields, from GATK and GROMACS to Quantum ESPRESSO and QGIS
• An always-on workstation that keeps your files between tasks, with a full Linux desktop for GUI tools like Fiji
• Every search, command and file edit on record. Replay the whole run any time
🧭 Claude and GPT, working for days
• A harness built for long-horizon research keeps one task on track from question to paper
• Cluster mode: a lead agent splits the job and hands each part to the model that does it best
• Self-review: before you see the paper, it rereads the draft against the sources and the code, and fixes what it finds
📄 Ends with a cited paper
• A LaTeX workbench with source and PDF side by side, compiled in the cloud
• Figures you can mark up: circle what's wrong, ask for a fix, keep every version
📱 On every device: web, macOS, Windows, iOS and Android.
Who it's for: researchers whose work runs on code, tools and data, from genomes to galaxies. If you already use Claude Code or Codex for research, think of ClawsGO Science as the lab behind the agent.
Try it free at clawsgo.ai. You start with 1,500 credits, no card needed.
Three questions for you: what's the first research question you'd hand it? If you use Claude Code or Codex for research today, where does it get stuck? And which tool or dataset should be in the lab?
Thanks for checking us out. We're here all day and will answer every comment 🙏