home / mcp / agent knowledge mcp server

Agent Knowledge MCP Server

Provides Elasticsearch-powered search, document validation, and secure management for AI assistants.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "itshare4u-agentknowledgemcp": {
      "command": "uvx",
      "args": [
        "agent-knowledge-mcp"
      ]
    }
  }
}

Agent Knowledge MCP is a complete knowledge management server for AI assistants. It provides powerful document storage, fast Elasticsearch-powered search, strict document validation, and secure, configurable operation so your AI can search, organize, and validate knowledge with confidence.

How to use

You connect your MCP-enabled AI to Agent Knowledge MCP to search, index, validate, and organize documents. Use it to quickly find API references, generate document templates, relate related topics, or update the server configuration. Interact with the server through your MCP client by selecting actions such as searching documents, indexing new content with tags, validating document structure, and adjusting settings.

How to install

Prerequisites you need before installing this MCP server are a runtime capable of running MCP commands and access to a shell where you can execute commands.

# Install with uvx (recommended)
uvx agent-knowledge-mcp

If you prefer to run a local development instance, you can start the server directly after cloning and installing dependencies, then run the main server script.

# Development setup (local runtime)
git clone https://github.com/itshare4u/AgentKnowledgeMCP.git
cd AgentKnowledgeMCP
pip install -r requirements.txt
python3 src/main_server.py

Configure a client to connect to the MCP server by providing the standard MCP connection snippet in your client’s config. The recommended stdio command is shown below for a connected client.

{
  "mcpServers": {
    "agent_knowledge": {
      "command": "uvx",
      "args": ["agent-knowledge-mcp"]
    }
  }
}

Security & configuration

Agent Knowledge MCP emphasizes enterprise-grade security with sandboxed operations, strict schema validation, and audit trails. It runs locally with no cloud dependencies unless you explicitly configure remote endpoints.

{
  "security": {
    "log_all_operations": true
  },
  "document_validation": {
    "strict_schema_validation": true,
    "allow_extra_fields": false
  }
}

Usage notes

Common tasks you can perform with your AI include searching documents for specific topics, indexing documents with tags, creating templates for API documentation, finding related documents, updating configuration, and validating document structure.

Available tools

Elasticsearch

Search, index, and manage documents within the knowledge base.

Document Validation

Enforce schema-based validation to ensure documents follow a defined structure.

Configuration

Manage server configuration and operational settings.

Security

Sandboxed operations with access controls and audit trails.