Security proxy for LLMs and MCP that blocks prompt injection, SSRF, injection attacks and credential exfiltration.
接入配置
选择客户端，复制配置，再按项目文档补齐环境变量。
claude mcp add mastyf --env ALERT_WEBHOOK_URL='<ALERT_WEBHOOK_URL>' -- npx -y @mastyf_ai/server~/.codex/config.toml
[mcp_servers.mastyf]
command = "npx"
args = ["-y", "@mastyf_ai/server"]
env = { ALERT_WEBHOOK_URL = "<ALERT_WEBHOOK_URL>" }
# codex mcp add mastyf --env ALERT_WEBHOOK_URL='<ALERT_WEBHOOK_URL>' -- npx -y @mastyf_ai/server~/.cursor/mcp.json
{
"mcpServers": {
"mastyf": {
"command": "npx",
"args": [
"-y",
"@mastyf_ai/server"
],
"env": {
"ALERT_WEBHOOK_URL": "<ALERT_WEBHOOK_URL>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"mastyf": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@mastyf_ai/server"
],
"env": {
"ALERT_WEBHOOK_URL": "<ALERT_WEBHOOK_URL>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"mastyf": {
"command": "npx",
"args": [
"-y",
"@mastyf_ai/server"
],
"env": {
"ALERT_WEBHOOK_URL": "<ALERT_WEBHOOK_URL>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"mastyf": {
"command": "npx",
"args": [
"-y",
"@mastyf_ai/server"
],
"env": {
"ALERT_WEBHOOK_URL": "<ALERT_WEBHOOK_URL>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"mastyf": {
"command": "npx",
"args": [
"-y",
"@mastyf_ai/server"
],
"env": {
"ALERT_WEBHOOK_URL": "<ALERT_WEBHOOK_URL>"
}
}
}
}需要的环境变量
ALERT_WEBHOOK_URL
把占位值换成你自己的密钥；不要把密钥提交到代码仓库。