home / mcp / onecite mcp server
Provides MCP-enabled access to OneCite’s citation processing via an MCP server for searching, batching, and formatting citations.
Configuration
View docs{
"mcpServers": {
"hzacode-onecite": {
"command": "onecite-mcp",
"args": [],
"env": {
"YOUR_ENV": "value"
}
}
}
}You can run the OneCite MCP server to enable any AI assistant or automation client to search, format, and generate citations from a wide range of sources. This MCP server exposes core functions like querying citations, batch processing references, and searching academic literature, making it easy to integrate citation capabilities directly into your workflow.
Connect your MCP client to the OneCite MCP server to access three core functions: cite, batch_cite, and search. You can use a local stdio-based server or a configured HTTP endpoint if you have that setup. The MCP interface lets you generate a single formatted citation, process multiple references at once, or search for papers by keywords. Use the interactive or callback options to resolve ambiguities when multiple matches exist.
Prerequisites you need before installing: Python and a working Python package manager.
pip install oneciteSet up the MCP server client integration in your editor or automation environment using the provided configuration snippet. The following examples show how to register OneCite under a standard MCP client setup.
{
"mcpServers": {
"onecite": { "command": "onecite-mcp" }
}
}The MCP exposes the following functions that you can invoke from your AI assistant or automation layer.
Functions:
- cite: Generate a single citation (e.g., for a DOI or a text query).
- batch_cite: Process a list of references and return formatted entries.
- search: Look up papers by keywords or identifiers.Generate a single citation for a given input such as a DOI, arXiv ID, or text query.
Process multiple references and return a set of formatted entries in a chosen style.
Search academic literature by keywords to retrieve relevant papers and metadata.