MCP server

Keystone

io.github.Srikarmk/keystone

Checked 26 Oct 2026, 08:29 UTC · Connected to it and listed its tools

What an arXiv paper builds on, disputes and assumes, read verbatim from its LaTeX. No LLM.

Website Source

At a glance

Type

MCP server

Publisher

Not verified

Status

Working at last check

Tools

3

Token cost

About 586 tokens per request

Sets up in

Claude Code, Claude Desktop, Codex, Gemini CLI, Cursor, VS Code, Antigravity, ChatGPT, OpenAI API, Grok

Last checked

26 Oct 2026, 08:29 UTC

Connect it to your agent

Add to CursorAdd to VS CodeVS Code Insiders

Opens the app, which shows the server and asks you to confirm. For other apps, copy the setup below.

Claude Code

claude mcp add --transport http keystone https://keystone-research.vercel.app/api/mcp

Claude Desktop

Settings → Connectors → Add custom connector, and paste the URL.

Codex (config.toml)

[mcp_servers.keystone]
url = "https://keystone-research.vercel.app/api/mcp"

Gemini CLI

{"mcpServers":{"keystone":{"httpUrl":"https://keystone-research.vercel.app/api/mcp"}}}

Cursor

{"mcpServers":{"keystone":{"url":"https://keystone-research.vercel.app/api/mcp"}}}

VS Code

{"servers":{"keystone":{"type":"http","url":"https://keystone-research.vercel.app/api/mcp"}}}

Antigravity

agy mcp add keystone https://keystone-research.vercel.app/api/mcp

ChatGPT

Settings → Apps & Connectors → Advanced settings: turn on Developer mode, then Create a connector and paste the URL. Which plans can do this varies.

OpenAI API (Responses tools)

{"type":"mcp","server_label":"keystone","server_url":"https://keystone-research.vercel.app/api/mcp","require_approval":"always"}

Grok (xAI API tools)

{"type":"mcp","server_label":"keystone","server_url":"https://keystone-research.vercel.app/api/mcp"}

Tools

• list_disagreements ~140 tokens
  Every sentence in the library where a paper says prior work is wrong, or puts itself beside it to win. Small on purpose — disagreement is rare and concentrated, and the whole set is readable.
• read_paper ~201 tokens
  Everything Keystone reads out of one arXiv paper: what it adopts, extends, disputes and measures against, and the assumptions it states with a citation, its own evidence, or nothing at all. Works for any paper on arXiv that ships LaTeX source — one outside the library is parsed on demand, which take
• search_library ~245 tokens
  Search 2497 readings taken from 234 arXiv papers' LaTeX source: citation stances (adopts, extends, disputes, measures against) and stated assumptions. Matches on the quoted sentence, the cue phrase that decided it, and the work cited. Returns the sentence verbatim with a link to the exact line.

Browse

• GitHub tools
• GitHub and code hosting
• Calling other AI models
• Research papers

Related MCP servers

• Pith
  Reviews of arXiv papers for AI agents: verdicts, claims, flaws, compiled records, citation graphs.
• Arxiv
  Search arXiv, fetch paper metadata, and read full-text content.
• Web Reader & Search (Tanod)
  Web search for agents: page to Markdown, screenshots, news search, metadata, sitemaps and RDAP.
• BrunoSan ArXiv Intelligence
  AI research intelligence for AI, ML, NLP, vision and robotics with citations and entity graphs.
• Web Search & Data API
  Web search, scraping, Google Trends and data lookups. Paid per call in USDC on Base via x402.
• AI/ML Research Papers
  AI/ML research papers from arXiv, DBLP, and HuggingFace
• LiveDataLink
  294 public-data tools across 60 domains; source freshness varies. Free tier.
• inktomd MCP Server
  Convert files, URLs, and documents to clean, AI-ready Markdown via MCP.

Badge for your README

[![AptLynx](https://aptlynx.net/badge/io.github.Srikarmk/keystone.svg)](https://aptlynx.net/servers/io.github.Srikarmk/keystone)

Listing text is written by the maker; AptLynx lists tools and does not vouch for them. Find another tool · Browse by topic · For agents