Oyemi MCP Server

1# Oyemi MCP Server
2 
3MCP (Model Context Protocol) server for the Oyemi semantic lexicon. Provides deterministic word-to-code mapping and valence analysis for AI agents like Claude, ChatGPT, and Gemini.
4 
5## Features
6 
7- **Semantic Encoding**: Convert words to deterministic semantic codes
8- **Valence Analysis**: Analyze text sentiment using lexicon-based valence
9- **Semantic Similarity**: Measure how similar two words are
10- **Synonym/Antonym Lookup**: Find related words
11- **Zero Runtime Dependencies**: No external NLP libraries needed at runtime
12 
13## Installation
14 
15```bash
16pip install oyemi-mcp
17```
18 
19Or install from source:
20 
21```bash
22git clone https://github.com/Osseni94/oyemi-mcp
23cd oyemi-mcp
24pip install -e .
25```
26 
27## Configuration
28 
29### Claude Desktop
30 
31Add to your `claude_desktop_config.json`:
32 
33```json
34{
35  "mcpServers": {
36    "oyemi": {
37      "command": "oyemi-mcp"
38    }
39  }
40}
41```
42 
43### Claude Code
44 
45Add to your MCP settings:
46 
47```json
48{
49  "mcpServers": {
50    "oyemi": {
51      "command": "oyemi-mcp"
52    }
53  }
54}
55```
56 
57## Available Tools
58 
59### `encode_word`
60Encode a word to its semantic code.
61 
62```
63encode_word("happy")
64-> {
65    "word": "happy",
66    "code": "1023-00012-3-2-1",
67    "pos": "adjective",
68    "abstractness": "abstract",
69    "valence": "positive"
70}
71```
72 
73### `analyze_text`
74Analyze the valence/sentiment of text.
75 
76```
77analyze_text("I feel hopeful but anxious about the future")
78-> {
79    "valence_score": 0.0,
80    "sentiment": "neutral",
81    "positive_words": ["hopeful"],
82    "negative_words": ["anxious"],
83    ...
84}
85```
86 
87### `semantic_similarity`
88Compare two words semantically.
89 
90```
91semantic_similarity("happy", "joyful")
92-> {
93    "similarity": 0.85,
94    "relationship": "very similar"
95}
96```
97 
98### `find_synonyms`
99Find synonyms for a word.
100 
101```
102find_synonyms("happy")
103-> {
104    "synonyms": ["glad", "felicitous", "well-chosen"]
105}
106```
107 
108### `find_antonyms`
109Find antonyms for a word.
110 
111```
112find_antonyms("happy")
113-> {
114    "antonyms": ["unhappy"]
115}
116```
117 
118### `batch_encode`
119Encode multiple words at once.
120 
121```
122batch_encode(["happy", "sad", "neutral"])
123-> {
124    "results": [
125        {"word": "happy", "valence": "positive"},
126        {"word": "sad", "valence": "negative"},
127        {"word": "neutral", "valence": "neutral"}
128    ]
129}
130```
131 
132### `get_lexicon_info`
133Get information about the lexicon.
134 
135```
136get_lexicon_info()
137-> {
138    "name": "Oyemi",
139    "version": "3.2.0",
140    "word_count": 145014
141}
142```
143 
144## Code Format
145 
146Oyemi codes follow the format `HHHH-LLLLL-P-A-V`:
147 
148| Component | Description | Values |
149|-----------|-------------|--------|
150| HHHH | Semantic superclass | 4-digit category code |
151| LLLLL | Synset ID | 5-digit unique identifier |
152| P | Part of speech | 1=noun, 2=verb, 3=adj, 4=adv |
153| A | Abstractness | 0=concrete, 1=mixed, 2=abstract |
154| V | Valence | 0=neutral, 1=positive, 2=negative |
155 
156## Use Cases
157 
158- **AI Sentiment Analysis**: Let AI agents understand emotional tone
159- **Semantic Grounding**: Provide concrete valence scores instead of guessing
160- **Text Analysis**: Analyze documents, reviews, feedback
161- **Word Relationships**: Find synonyms, antonyms, similar words
162 
163## License
164 
165MIT License
166 
167## Author
168 
169Kaossara Osseni - [grandnasser.com](https://grandnasser.com)
170