paper-fetch-skill
About this tool
Turn a DOI, arXiv ID, URL or title into structured Markdown full text with YAML metadata, sections, equations, tables, figure captions and references. Available as a CLI, the paper-fetch-mcp server or an agent skill. It tries publishers' official HTML, XML or APIs first, then falls back to validated PDFs. Use it within your own or your institution's access rights.
- Feature 1: Support for around 20 publishers and sources, including Elsevier, Springer, Wiley, IEEE, ACS, PLOS, Frontiers and arXiv.
- Feature 2: Markdown output retains sections, equations, tables, captions and references.
- Feature 3: Download paper images and link to local copies.
- Feature 4: Batch retrieval and caching.
- Feature 5: The README states that it does not bypass paywalls, CAPTCHAs or institutional access controls.
Resource information
- Maintainer
- Independent developer Dictation354.
- Type
- MCP
- Category
- Literature search & access
- Runs on
- MCP
- Who it's for
- Researchers reading papers closely, writing reviews or extracting data, using their own or institutional subscriptions to prepare agent-readable text.
- License at time of listing
- MIT
- Source
- GitHub
