MCP serversx402 APIsAI dev toolsFree APIsSearchSubmit a toolGet listed / Advertise
Home › MCP servers › mlava/scholar-sidekick-mcp

mlava/scholar-sidekick-mcp

Resolve scholarly identifiers (DOI, PMID, ISBN, arXiv and more) to CSL JSON, format citations in thousands of styles, or export to BibTeX, RIS or EndNote.

Visit github.com

CategoryResearch
Languagetypescript
Runscloud
OSlinux, macos, windows
OfficialCommunity

Source: punkpeye/awesome-mcp-servers · Licence: MIT. Listed in MCP servers. We list tools; we do not test, rate or endorse them.

Is this your tool? Update the description or add details for free via Submit / claim. Want it at the top of MCP servers? See Get listed / Advertise (paid placements are labelled Sponsored).

Related

HasData/google-trends-mcp

Google Trends data as JSON: interest over time and by region, plus rising and top related queries and topics for any keyword.

HasData/google-scholar-mcp

Google Scholar search as JSON with author and year filters, cited-by and all-versions lookups, and formatted citations with BibTeX export.

Agnuxo1/benchclaw-integrations

Register LLMs and agents and submit Markdown research papers to the BenchClaw leaderboard, where a panel of judges scores them with deception…

amos689/paper-preflight

Checks every reference of a LaTeX paper against Crossref, dblp, arXiv, DataCite, PubMed and OpenAlex records: whether each cited work exists,…

ankitkapur1992-hlido/hlido-mcp

Hlido trust scores, claim audits and comparisons for reviewed AI agents: tier verdicts, per-claim pass/fail audits and signed evidence.

archoor/painspotter-mcp

Startup and product opportunities mined from Reddit, Hacker News, Product Hunt and Stack Exchange: pain points, trending themes and commercial scores.

Battam1111/omniseek

Self-hosted perception server: local bilingual ASR, vision, cross-lingual search, credentialed login-walled sources and a persistent relation graph.

BrowseAI-HQ/BrowserAI-Dev

Evidence-backed web research: real-time search with cited claims, confidence scores and a mode comparing raw LLM answers with evidence-backed ones.