A composable safety middleware that gives any AI application content moderation, PII protection, injection defense, and audit logging in a single YAML configuration.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add guardianshield -- uvx guardianshield~/.codex/config.toml
[mcp_servers.guardianshield]
command = "uvx"
args = ["guardianshield"]
# codex mcp add guardianshield -- uvx guardianshield~/.cursor/mcp.json
{
"mcpServers": {
"guardianshield": {
"command": "uvx",
"args": [
"guardianshield"
]
}
}
}.vscode/mcp.json
{
"servers": {
"guardianshield": {
"type": "stdio",
"command": "uvx",
"args": [
"guardianshield"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"guardianshield": {
"command": "uvx",
"args": [
"guardianshield"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"guardianshield": {
"command": "uvx",
"args": [
"guardianshield"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"guardianshield": {
"command": "uvx",
"args": [
"guardianshield"
]
}
}
}