home / mcp / inkeep mcp server
Inkeep MCP Server
Configuration
View docs{
"mcpServers": {
"inkeep-mcp-server-python": {
"command": "uv",
"args": [
"--directory",
"<YOUR_INKEEP_MCP_SERVER_ABSOLUTE_PATH>",
"run",
"-m",
"inkeep_mcp_server"
],
"env": {
"INKEEP_API_KEY": "<YOUR_INKEEP_API_KEY>",
"INKEEP_API_MODEL": "inkeep-rag",
"INKEEP_API_BASE_URL": "https://api.inkeep.com/v1",
"INKEEP_MCP_TOOL_NAME": "search-product-content",
"INKEEP_MCP_TOOL_DESCRIPTION": "Retrieves product documentation about Inkeep. The query should be framed as a conversational question about Inkeep."
}
}
}
}You deploy this MCP server to power retrieval-augmented conversations using your product content. It runs locally via uv and connects to Inkeep to fetch and index your data, letting your MCP client query up-to-date information through a consistent interface.
To use this MCP server, configure your MCP client to connect to the Inkeep MCP endpoint and supply your API credentials. You will run the server locally and point your client to the stdio-based MCP entry that starts the uv runner with the inkeep module. The server exposes a retrieval tool that your client can query to fetch product documentation and related data.
Prerequisites: You need a Python-enabled environment with the uv workflow manager installed. You also need an Inkeep account to manage and provide the RAG and an API key for authentication.
Step 1: Clone the project.
git clone https://github.com/inkeep/mcp-server-python.git
cd mcp-server-python
uv venv
uv pip install -r pyproject.tomlNote the full path of the project directory. This path is referenced as <YOUR_INKEEP_MCP_SERVER_ABSOLUTE_PATH> in the configuration example.
Get an API key from Inkeep to authenticate requests. You will use this key in your MCP client configuration.
Add the MCP server configuration to your client as shown in the example. This configuration runs the uv command to start the Inkeep MCP server module and passes the necessary environment variables.
The full configuration snippet is provided below and should be placed in your client’s MCP server list under the mcpServers entry. Ensure the path and API key placeholders are replaced with your actual values.
json claude_desktop_config.json
{
"mcpServers": {
"inkeep-mcp-server": {
"command": "uv",
"args": [
"--directory",
"<YOUR_INKEEP_MCP_SERVER_ABSOLUTE_PATH>",
"run",
"-m",
"inkeep_mcp_server"
],
"env": {
"INKEEP_API_BASE_URL": "https://api.inkeep.com/v1",
"INKEEP_API_KEY": "<YOUR_INKEEP_API_KEY>",
"INKEEP_API_MODEL": "inkeep-rag",
"INKEEP_MCP_TOOL_NAME": "search-product-content",
"INKEEP_MCP_TOOL_DESCRIPTION": "Retrieves product documentation about Inkeep. The query should be framed as a conversational question about Inkeep."
}
},
}
}