MHLabs MCP Tools Server

You can run and connect to the Modular MCP Tools Server to load, manage, and serve AI tools through the MCP protocol. It supports multiple transports (STDIO and HTTP) and exposes a growing ecosystem of NLP and text preprocessing components that you can access from MCP clients to perform tasks like tokenization, lemmatization, normalization, and more. This guide walks you through practical usage, installation steps, and essential configuration to get you up and running quickly.

How to use

Start the MCP server in STDIO mode for local tool access or in HTTP mode for web-based clients. In STDIO mode you can run the server directly from your development environment and connect with a client that communicates via standard input/output. In HTTP mode, the server binds a web endpoint so MCP clients can reach it over the network.

How to install

Prerequisites you need before installation: a Python 3.x runtime and a working Python package manager (pip). You should also have a code editor for development and testing.

Step-by-step installation and setup you can follow locally.

Additional notes and usage considerations

Configuration is managed through environment variables and command-line options. You can enable or disable authentication, choose the transport, and set server details. For development, you can run in debug mode to capture detailed logs.

Available tools

tokenize

Text tokenization component that splits text into discrete tokens for downstream processing.

pos

Part-of-speech tagging to identify grammatical categories for tokens.

lemma

Lemmatization to reduce words to their base forms.

morphology

Morphology analysis to study word forms and inflections.

dep

Dependency parsing to reveal syntactic relationships between tokens.

ner

Named Entity Recognition to identify and classify entities in text.

norm

Text normalization to standardize text input.

to_lower

Convert text to lowercase to ensure uniformity.

to_upper

Convert text to uppercase for emphasis or normalization.

remove_number

Remove numeric characters from text.

remove_url

Strip URLs from text passages.

remove_punctuation

Eliminate punctuation marks from text.

remove_stopword

Remove common stopwords to reduce noise.

remove_html

Remove HTML tags from text.

expand_contraction

Expand contracted forms (e.g., can't -> cannot) for clarity.

tokenize_word

Tokenize individual words.

tokenize_sentence

Tokenize text into sentences.

stem_word

Stem words to their root forms.

lemmatize_word

Lemmatize words to their base lemmas.

preprocess_text

Combine multiple preprocessing steps into a single workflow.