Wolfram Language Paclet Repository
Community-contributed installable additions to the Wolfram Language
Client for a Python-made MCP server
Introduction
This notebook has a complete usage example of a thin Model Context Protocol (MCP) client of the Raku package "MCPClient".
The MCP server is created and run in Python -- see the file . That MCP server provides the methods add , random_words_list , random_pet_names_list , and random_pretentious_job_title_list . The corresponding Python packages have to be installed -- see the from ... import ... statements in that Python file.
"MCPClient" provides a set of functions for using MCP server's methods by converting them to objects. Those objects can be used with Wolfram Language's Large Language Model (LLM) framework. (, , .)
Install the paclet and load it
In[76]:=
PacletInstall["AntonAntonov/MCPClient"];Needs["AntonAntonov`MCPClient`"]
Setup and initialization
Setup MCP server process creation command elements:
In[78]:=
cmd={FileNameJoin[{$HomeDirectory,"/miniforge3/bin/python"}],"-i",FileNameJoin[{PacletObject["AntonAntonov/MCPClient"]["Location"],"Resources","simple_mcp_server.py"}]};
Create the MCP client object:
In[79]:=
client=
[cmd];
 | MCPStart |
Function[ |
Out[80]=
Initialize the client:
In[81]:=
r=
[client];
| MCPInitialize |
Function[ |
Out[82]=
Tools discovery and creation
Get the MCP server tools list (and tabulate parts of it):
In[94]:=
mcpTools=
[client];
| MCPListTools |
Function[ |
Out[95]=
Make LLMTool objects:
In[85]:=
tools=
[client,mcpTools];
| MCPToLLMTool |
Function[ |
Out[86]=
Make a configuration with the tools:
In[87]:=
conf=LLMConfiguration["Model"{"OpenAI","gpt-5.3-chat-latest"},"MaxTokens"8192,"Tools"tools];
LLM invocations
Find the unaccepted GitHub Pull Requests (PRs):
In[88]:=
LLMSynthesize["Give the current time and generate a list of 12 random cat pet names.",LLMEvaluatorconf]
Out[88]=
Current time: 2026-05-18 12:01:40Here are 12 random cat names:Tofu, Gracie, James, Olive, Katharine, Greyson, Elsa, Max, Creep, Clay, Korra, Winston
Generate a JSON object of random pretentious jobs, and deserialize and print it:
In[89]:=
fJobs=LLMFunction[{"Generate a list of `1` random pompous-sounding jobs, in English, and `2` in Bulgarian.",LLMPrompt["NothingElse"]["JSON"]},LLMEvaluatorconf];
In[90]:=
res=fJobs[6,3];res=ImportString[ToString[res,CharacterEncoding"UTF-8"],"JSON"]
Out[91]=
english{Future Applications Facilitator,District Interactions Facilitator,Global Web Engineer,Future Paradigm Coordinator,District Identity Assistant,Customer Response Strategist},bulgarianКорпоративен Плановик на Сигурност,Динамичен Оркестратор на Изследвания,Главен Продуцент по Интеграция
 |
|
 |
|
""