Oyemi MCP

🚀 Oyemi MCP Server

The Oyemi MCP (Model Context Protocol) server is designed for the Oyemi semantic lexicon. It offers deterministic word - to - code mapping and valence analysis for AI agents such as Claude, ChatGPT, and Gemini.

✨ Features

• Semantic Encoding: Transform words into deterministic semantic codes.
• Valence Analysis: Evaluate text sentiment using lexicon - based valence.
• Semantic Similarity: Quantify the similarity between two words.
• Synonym/Antonym Lookup: Locate related words.
• Zero Runtime Dependencies: No external NLP libraries are required at runtime.

📦 Installation

You can install the oyemi - mcp package using pip:

pip install oyemi-mcp

Or install it from the source:

git clone https://github.com/Osseni94/oyemi-mcp
cd oyemi-mcp
pip install -e .

📚 Documentation

Configuration

Claude Desktop

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "oyemi": {
      "command": "oyemi-mcp"
    }
  }
}

Claude Code

Add the following to your MCP settings:

{
  "mcpServers": {
    "oyemi": {
      "command": "oyemi-mcp"
    }
  }
}

💻 Usage Examples

Basic Usage

Here are some basic usage examples of the available tools:

encode_word

Encode a word to its semantic code.

encode_word("happy")
-> {
    "word": "happy",
    "code": "1023-00012-3-2-1",
    "pos": "adjective",
    "abstractness": "abstract",
    "valence": "positive"
}

analyze_text

Analyze the valence/sentiment of text.

analyze_text("I feel hopeful but anxious about the future")
-> {
    "valence_score": 0.0,
    "sentiment": "neutral",
    "positive_words": ["hopeful"],
    "negative_words": ["anxious"],
    ...
}

semantic_similarity

Compare two words semantically.

semantic_similarity("happy", "joyful")
-> {
    "similarity": 0.85,
    "relationship": "very similar"
}

find_synonyms

Find synonyms for a word.

find_synonyms("happy")
-> {
    "synonyms": ["glad", "felicitous", "well-chosen"]
}

find_antonyms

Find antonyms for a word.

find_antonyms("happy")
-> {
    "antonyms": ["unhappy"]
}

batch_encode

Encode multiple words at once.

batch_encode(["happy", "sad", "neutral"])
-> {
    "results": [
        {"word": "happy", "valence": "positive"},
        {"word": "sad", "valence": "negative"},
        {"word": "neutral", "valence": "neutral"}
    ]
}

get_lexicon_info

Get information about the lexicon.

get_lexicon_info()
-> {
    "name": "Oyemi",
    "version": "3.2.0",
    "word_count": 145014
}

🔧 Technical Details

Code Format

Oyemi codes follow the format HHHH-LLLLL-P-A-V:

📚 Use Cases

• AI Sentiment Analysis: Enable AI agents to comprehend emotional tone.
• Semantic Grounding: Offer concrete valence scores instead of guesswork.
• Text Analysis: Analyze documents, reviews, and feedback.
• Word Relationships: Discover synonyms, antonyms, and similar words.

📄 License

This project is under the MIT License.

👨‍💻 Author

Kaossara Osseni - grandnasser.com