AI Clients Setup
Configuration in AI clients
Integration with Main AI Clients
Claude Desktop
Add to claude_desktop_config.json:
The –config arguments are only necessary if you want to use a custom configuration file. By default the installation through the installation script creates the configuration file in the default path and this parameter is not necessary.
{
"mcpServers": {
"remembrances": {
"command": "/path/to/remembrances-mcp",
"args": [
"--config",
"/path/to/config.yaml"
]
}
}
}
Simpler:
{
"mcpServers": {
"remembrances": {
"command": "/path/to/remembrances-mcp",
"args": []
}
}
}
Github Copilot
Add in your mcp.json
{
"servers": {
"remembrances": {
"command": "/path/to/remembrances-mcp",
"args": []
}
}
}
MCP Client in Python
from mcp import ClientSession, StdioServerParameters
from mcp.client.stdio import stdio_client
server_params = StdioServerParameters(
command="/path/to/remembrances-mcp",
args=["--config", "/path/to/config.yaml"]
)
async with stdio_client(server_params) as (read, write):
async with ClientSession(read, write) as session:
await session.initialize()
# Save a fact
result = await session.call_tool(
"remembrance_save_fact",
arguments={"key": "test", "value": "Hello"}
)
# Search
results = await session.call_tool(
"remembrance_search",
arguments={"query": "test", "limit": 5}
)