Jeanibarz Knowledge Base MCP Server

🚀 Knowledge Base MCP Server

This MCP server provides you with tools to list and retrieve content from different knowledge bases.

🚀 Quick Start

📦 Installation

These instructions assume that you have already installed Node.js and npm on your system.

Prerequisites

• Node.js (version 16 or higher)
• npm (Node Package Manager)

1. Clone the repository:

git clone <repository_url>
cd knowledge-base-mcp-server

2. Install dependencies:

npm install

3. Configure environment variables:

• The server requires the HUGGINGFACE_API_KEY environment variable to be set. This is the Hugging Face Inference API key used to generate embeddings for knowledge base content. You can obtain an API key for free on the Hugging Face website (https://huggingface.co/).
• The server requires the KNOWLEDGE_BASES_ROOT_DIR environment variable to be set. This variable specifies the directory where the knowledge base subdirectories are located. If not set, it will default to $HOME/knowledge_bases, where $HOME is the current user's home directory.
• The server supports the FAISS_INDEX_PATH environment variable to specify the path to the FAISS index. If not set, it defaults to $HOME/knowledge_bases/.faiss.
• The server supports the HUGGINGFACE_MODEL_NAME environment variable to specify the Hugging Face model used to generate embeddings. If not set, it defaults to sentence-transformers/all-MiniLM-L6-v2.
• You can configure these environment variables in the .env file or in the system environment.

4. Build the server:

npm run build

5. Add to MCP settings: To add this server to the MCP interface, follow the instructions in the MCP documentation. Ensure that the MCP is correctly configured to recognize this server and set the necessary permissions.

6. Create knowledge base directories: Before you start, it is recommended that you create an appropriate directory structure for each knowledge base. For example:

knowledge_bases/
  ├── doc1.md
  ├── doc2.pdf
  └── subdirectory/
      └── file.txt

💻 Usage Examples

Features of the Knowledge Base Server

This server provides the following main functions:

• List knowledge bases: Retrieve all registered knowledge bases and their metadata.
• Retrieve knowledge content: Retrieve relevant knowledge documents based on a query statement. It supports searching within a specific knowledge base or across all knowledge bases.

Knowledge Retrieval Tool

The retrieve_knowledge tool performs semantic searches using a FAISS index. This index is automatically updated when the server starts or when changes to knowledge base files are detected.

Output Format

The output of the retrieval operation is a Markdown-formatted string with the following structure:

## Semantic Search Results

**Result 1:**

[Content of the most similar snippet]

**Source:**
```json
{
  "source": "[Path to the file containing the snippet]"
}

Result 2:

[Content of the second most similar snippet]

Source:

{
  "source": "[Path to the file containing the snippet]"
}

Disclaimer: The provided results may not all be relevant. Please check the relevance of the information.

Each result includes the content of the most similar snippet, the path to the source file, and a similarity score.

### Example Usage
Using the tool through the MCP interface:
```json
{
  "tool": "retrieve_knowledge",
  "args": {
    "query": "Your query statement.",
    "knowledge_base": "Specify the knowledge base name (optional)",
    "threshold": 2.0
  }
}

Parameters

• query: Required. The query statement for the information to be retrieved.
• knowledge_base: Optional. Specify a specific knowledge base for the search; if not specified, it defaults to searching all knowledge bases.
• threshold: Optional. The similarity threshold below which results will be returned. The default is 2.0.

Example Output

{
  "result": [
    {
      "content": "Relevant snippet content.",
      "source": "File path",
      "score": 1.5
    },
    {
      "content": "Another relevant snippet content.",
      "source": "Another file path",
      "score": 1.8
    }
  ],
  "message": "Search completed."
}

⚠️ Important Note

• Ensure that all knowledge base files are readable and that the server has appropriate file permissions.
• Regularly update the knowledge bases to keep the index up-to-date.
• Adjust the similarity threshold to optimize the relevance of the retrieval results.

By following these steps, you can effectively utilize this MCP server to manage, search, and retrieve your knowledge content.