citewire
Search preprint metadata across physics, computer science, mathematics, and more.
Search computer science bibliography.
Search papers and author graph.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@citewireSearch for recent research on protein folding"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
citewire
An attribution-first MCP server for free news sources and free article APIs.
Why it exists
Agents need to read news and research metadata, but the publishers who create that work keep the rights and the traffic. citewire reads the metadata and links straight back to the original article, so an agent can list, search, and cite a story without the content ever being copied, stored, or republished. The name is the whole idea: cite plus wire.
Related MCP server: Academic Paper MCP HTTP/SSE Server
What it does
citewire has two jobs.
Serve a news platform. Any site that exposes a compatible read API (list, detail, topics) becomes a set of typed MCP tools (
news.list,news.get,news.topics,news.about) by declaring one small JSON block. The tools are a mechanical projection of the HTTP surface, so they cannot drift from the API they wrap. See docs/platform-contract.md.Query free article APIs. Ten open news and research APIs are available as pass-through tools. Nothing is stored and nothing is republished; each tool returns metadata and links to the original. Every provider is disabled by default. Enabling one is your deliberate act, and reviewing that provider's terms for your use is your responsibility. See docs/providers.md.
Quick start
Run over stdio (the default transport, and how MCP clients launch the server):
npx citewire --config citewire.config.jsonWire it into Claude Desktop (claude_desktop_config.json) or Claude Code
(.mcp.json):
{
"mcpServers": {
"citewire": {
"command": "npx",
"args": ["citewire", "--config", "citewire.config.json"]
}
}
}Use an absolute path for --config if the client does not launch from your
project directory.
Serve over Streamable HTTP instead of stdio (default port 8722, JSON-RPC over
POST /):
npx citewire --config citewire.config.json --http 8722Configuration
A deployment is described entirely by a small JSON file: an optional platform
to wrap, and the providers you choose to enable. This example wraps the Karaya
Group Industry News platform and enables one provider, OpenAlex:
{
"platform": {
"name": "Karaya Group Industry News",
"siteUrl": "https://karaya.group",
"apiBase": "https://karaya.group/api/v1/news"
},
"providers": {
"openalex": { "enabled": true }
}
}Omit the platform block to run providers only. Omit providers, or leave a
provider out of it, and that provider stays disabled. A bad config fails loudly
at startup with an error that names the offending field.
Platform tools
These four tools exist only when a platform block is configured.
Tool | Purpose |
| List items newest first. Filters: |
| One item by |
| The active taxonomy topics (slug, label, description). |
| Static facts about the server and platform, including the attribution policy. |
Providers
Ten free APIs. None requires an API key. Every one ships disabled. Provider
tools are namespaced under a short prefix; the authoritative list for a given
deployment is whatever the running server advertises through tools/list. Full
detail, endpoints, and free-access terms are in
docs/providers.md.
Provider | Tool | Queries | Key required | Default |
GDELT DOC 2.0 |
| Worldwide news article metadata | none | disabled |
GDELT Context 2.0 |
| Snippet-level context around a search term | none | disabled |
arXiv |
| Preprint metadata across physics, CS, math, and more | none | disabled |
OpenAlex |
| Scholarly works, authors, and venues | none | disabled |
Crossref |
| DOI registration metadata | none | disabled |
Semantic Scholar |
| Papers and author graph | none | disabled |
Europe PMC |
| Life-sciences and biomedical literature | none | disabled |
dblp |
| Computer-science bibliography | none | disabled |
Hacker News |
| HN stories and items (Firebase API) | none | disabled |
DEV |
| Published DEV articles | none | disabled |
Design
Zero runtime dependencies. ESM, Node >= 18. Nothing to audit but the source in this repo.
Hand-written JSON-RPC. The MCP core is a small dispatcher, not a wrapped SDK.
Stateless. No sessions and no stored state. Each request is handled on its own.
Transport-agnostic core.
createCitewire(config)builds one server; stdio and Streamable HTTP both pump messages through it. The HTTP handler is plain Node and also mounts inside a serverless function.Storage is out of scope. citewire never stores or republishes content. It reads metadata and links to the original. That boundary is the point of the project, not a limitation of it.
Deployments
Karaya Group's Industry News platform (karaya.group/industry-news) is the first production deployment. It serves citewire at https://karaya.group/mcp.
Contributing
See CONTRIBUTING.md. Tests run with npm test (node --test,
no install required).
License
MIT. See LICENSE.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseAqualityDmaintenanceMCP server enabling AI agents to search and retrieve scientific papers, citations, and author profiles from Crossref, OpenAlex, and Semantic Scholar with no API keys required.Last updated53MIT
- Alicense-qualityDmaintenanceA MCP server for academic literature retrieval, aggregating multiple data sources like arXiv, Crossref, OpenAlex, PubMed, and Semantic Scholar to provide search, details, citations, trends, and recommendations.Last updated4MIT
- Alicense-qualityAmaintenanceZero-auth multi-source research MCP server that enables web search, reading URLs, PDFs, GitHub repos, and querying Hacker News, Stack Overflow, Semantic Scholar, and YouTube transcripts without API keys.Last updated10Apache 2.0
- Alicense-qualityAmaintenanceSearch, trending, topics, and local news all in one MCP server. Article previews, deduplication, source filtering, and 40+ languages built in.Last updated492MIT
Related MCP Connectors
MCP server for Altmetric APIs - track research attention across news, policy, social media, and more
Academic research MCP server for paper search, citation checks, graphs, and deep research.
Citable retrieval across papers, books, patents, Wikipedia, and live social sources.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/MeekPhills/citewire'
If you have feedback or need assistance with the MCP directory API, please join our Discord server