Curated by Vinod Kumar Jain & Amit Jain · All Frontier Global · free, no login
Give a topic and this does two separate things: drafts thesis candidates sharp enough to argue with, and searches Crossref for real papers on the same subject. The first is generated and the second is retrieved, and the page never mixes them up. Submitting AI-written work as your own may violate your institution’s rules.
Thesis candidates (AI-drafted)
Papers on this topic (real Crossref results)
The two halves, and why they are kept apart
A drafting and structuring aid, not a submission. Submitting AI-written work as your own may violate your institution’s rules, and increasingly does. Use this to outline, to argue against, and to improve your own writing — not to replace it. If you are unsure what your institution permits, ask before you use it, not after.
The thesis candidates are generated. They are shapes for an argument — a claim specific enough to be wrong. Tighten or expand them, then write your own.
The papers are real. They come from Crossref’s public API through this site’s own proxy, and every title, author and DOI is a record a publisher deposited. Verify relevance by reading them — a keyword match is not a reading, and Crossref’s ranking is not a judgement about your argument.
The model never sees the papers and the papers never inform the model. That separation is deliberate: it is what stops a generated thesis from arriving wearing a citation it did not earn.
No citation here is generated. If Crossref returns nothing, the page says so rather than filling the space.
This is one of six tools on the estate that use the network. It talks only to this site’s own endpoints — nothing you type goes to a third party directly, and nothing is stored. The other fifty-seven run entirely in your browser.
Honest limits
A thesis you did not arrive at is hard to defend. Use these to see what a sharp claim looks like, then find your own — in a viva the difference shows immediately.
Crossref search is keyword relevance, not a systematic review, and coverage varies by field and publisher. For anything serious, search a proper subject index as well.
Nothing here checks retraction status, quality or replication. A paper appearing in the list is a record that exists, nothing more.
Nothing you type is stored, and the workshop disappears when you close the tab.
Sources
Papers come from Crossref’s public REST API, fetched at request time through this site’s own same-origin proxy and never cached here; links resolve through doi.org. The thesis candidates come from a language model and carry no citations, because they have none. Your institution’s academic-integrity policy is the authority on what you may submit.
One request form is used and only one: https://api.crossref.org/works?query= followed by your topic, sent through this site’s own proxy so your browser never contacts Crossref itself. Outbound paper links are built as https://doi.org/ plus the DOI and nothing else. The language model is never given a paper and is never asked for one.