Scriptora AI Docs
References

Searching References

Search 6 academic databases simultaneously to find papers and add them to your bibliography — manually or via the AI pipeline.

Scriptora AI offers two ways to find and add references: the References sidebar for manual search, and the AI pipeline which runs automated multi-database searches as part of Stage 3 of the creation workflow.

Literature search results streaming from multiple databases
The AI pipeline searches arXiv, OpenAlex, PubMed, Semantic Scholar, CrossRef, and CORE simultaneously, streaming results in real time.

Manual search (References sidebar)

Click the References tab in the left sidebar (bookmark icon).

Enter a paper title, author name, keyword phrase, or DOI. Scriptora queries all six databases in parallel.

Each result shows the title, authors, year, source badge, citation count, and open-access indicator.

Click + Add to .bib next to any result. The BibTeX entry is appended to your main .bib file immediately.

Database coverage

BadgeDatabaseCoverage
arXivarXiv.orgPreprints in CS, physics, math, biology
OAOpenAlex250M+ scholarly works, open metadata
PubMedMEDLINE / PubMedBiomedical and life sciences
S2Semantic ScholarAI-powered, 200M+ papers
CRCrossRefDOI registry, journal articles
CORECORE.ac.ukOpen access full text

Results are deduplicated automatically — if the same paper appears in multiple databases, you see it once.

Adding a reference by DOI or arXiv ID

  1. Click the BookPlus icon in the sidebar toolbar (or ask the AI: "Add the paper with DOI 10.1038/s41586-021-03819-2")
  2. Paste a DOI like 10.1145/3290605.3300362 or an arXiv ID like 2301.07041
  3. Scriptora fetches metadata from CrossRef or arXiv API
  4. Click Add to .bib — the entry is inserted as a properly formatted @article or @misc block

Sorting and filtering results

Use the sort control at the top of the results list:

Sort optionDescription
RelevanceDefault — AI-ranked by semantic match to your query
Citation countMost cited first
YearNewest first

AI-assisted literature search (pipeline Stage 3)

When running the 7-stage creation pipeline, Stage 3 performs a fully automated literature search:

  1. The AI generates 3–5 targeted search queries from your research question
  2. Each query runs across all 6 databases in parallel
  3. Results stream into the chat panel grouped by query
  4. You review the results and select the papers to include (typically 8–15)
  5. Click Continue with N selected papers to advance to Stage 4

The selected papers are automatically added to references.bib and used as the evidence base for all subsequent writing stages.

Deselect papers that are only tangentially related to your specific question. The AI writes more focused content when given a curated, relevant set rather than every result.

You can also trigger a literature search directly from the chat:

Search for the top 5 recent papers on contrastive self-supervised learning
and add the best ones to my references.

The AI will search, display results in the chat, and — with your approval — append BibTeX entries to references.bib and insert \cite{} commands in the relevant section.

On this page