GUARDRAIL is a security framework designed to protect Large Language Model (LLM) applications using the Model Context Protocol (MCP). It focuses on preventing data exfiltration and unauthorized access while ensuring resource security.
GUARDRAIL is a security framework designed to protect Large Language Model (LLM) applications using the Model Context Protocol (MCP). It focuses on preventing data exfiltration and unauthorized access while ensuring resource security.
Add this configuration to your claude_desktop_config.json:
{
"mcpServers": {
"guardrail": {
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-guardrail"
]
}
}
}Restart Claude Desktop, then ask:
"What tools do you have available from GUARDRAIL?"
No configuration required. This server works out of the box.
"What resources are available in GUARDRAIL?"
Claude will query available resources and return a list of what you can access.
"Show me details about [specific item] in GUARDRAIL"
Claude will fetch and display detailed information about the requested item.
"Create a new [item] in GUARDRAIL with [details]"
Claude will use the appropriate tool to create the resource and confirm success.
We build custom MCP integrations for B2B companies. From simple connections to complex multi-tool setups.