federicodeponte/opendraft

Write research paper and literature review drafts with an open-source Python engine that checks citation DOIs against scholarly databases. Export PDF, Word, or LaTeX.

OpenDraft

What it is – An open‑source Python engine that lets an AI agent generate full‑length academic drafts (papers, literature reviews, thesis chapters) with citations that are automatically verified against public scholarly databases (CrossRef, OpenAlex, Semantic Scholar).

Why it matters – It tackles two big problems in AI‑generated research: (1) hallucinated references and (2) the difficulty of producing long‑form, structured documents. By splitting the workflow into 19 specialised agents and only keeping sources whose DOI appears in at least two databases, the tool gives researchers a reproducible, source‑grounded first draft that can be edited and polished further.

Key features

  • 19 dedicated AI agents covering research, outlining, writing, citation checking, polishing and export.
  • Citation verification: a DOI must be present in ≥2 of CrossRef, OpenAlex, Semantic Scholar (configurable).
  • Supports 57+ languages and can export to PDF, Word (.docx) or LaTeX.
  • CLI and Python library; can be run locally with your own LLM API keys (Gemini Flash, Claude Opus, etc.).
  • Additional utilities: TL‑DR 5‑bullet summaries, 60‑second audio digests (ElevenLabs TTS), data‑fetching from World Bank, Eurostat, OWID, and draft revision/versioning.
  • Can be used as a Claude/Edge skill (getedge.cc/skills/autonomous‑research).

Typical workflow

  1. Provide a topic line (e.g., Write a paper on Neural Networks in Healthcare).
  2. The research agents query CrossRef, OpenAlex and Semantic Scholar, keep only multi‑source‑confirmed DOIs, and build a bibliography.
  3. Structure agents create an outline; writing agents draft each section.
  4. Citation‑claim verifier checks relevance of each source.
  5. Polish agents refine language; export agents produce PDF/Word/LaTeX.
  6. Optionally run TL‑DR, digest, or revision commands.

Who should use it – Academic researchers, graduate students, open‑source maintainers building reproducible research pipelines, and developers who want a free, inspectable alternative to proprietary AI writing tools.

Limitations – The tool does not replace human expertise; citations are only existence‑checked, not guaranteed to support every claim. It requires an LLM API key and works best with models like Gemini Flash or Claude Opus. It is not intended for cheating or producing final publishable manuscripts without human review.

Links

Related

  • Project
    CYC2002tommy/Deep-Research-Agent

    An autonomous AI agent pipeline for rigorous academic research that automates literature review, full-text verification, and APA 7th document generation.

  • Project
    hoolulu/deep-research

    An automated multi-agent system that generates institutional-grade, citable research reports in 19 languages using a four-stage search and verification pipeline.

  • Project
    Tswoen/Paper-Agent

    An AI-powered research assistant that automates the process of searching, reading, and analyzing academic papers to generate comprehensive, evidence-based literature reviews.

  • Project
    wp-a/nature-academic-search

    A literature retrieval and citation verification tool that integrates multiple academic databases to prevent AI hallucinations and export verified references.

  • Project
    gianlucasb/hallucinator

    A tool that detects hallucinated citations in academic papers by extracting references from PDFs and validating them against multiple academic databases.