MCP security trust layer. Scan packages, get trust verdicts, check incident exposure, query abuse database. 5 tools, no API key needed.
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 agentscore -- npx -y @agentscore-xyz/mcp-server~/.codex/config.toml
[mcp_servers.agentscore]
command = "npx"
args = ["-y", "@agentscore-xyz/mcp-server"]
# codex mcp add agentscore -- npx -y @agentscore-xyz/mcp-server~/.cursor/mcp.json
{
"mcpServers": {
"agentscore": {
"command": "npx",
"args": [
"-y",
"@agentscore-xyz/mcp-server"
]
}
}
}.vscode/mcp.json
{
"servers": {
"agentscore": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@agentscore-xyz/mcp-server"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"agentscore": {
"command": "npx",
"args": [
"-y",
"@agentscore-xyz/mcp-server"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"agentscore": {
"command": "npx",
"args": [
"-y",
"@agentscore-xyz/mcp-server"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"agentscore": {
"command": "npx",
"args": [
"-y",
"@agentscore-xyz/mcp-server"
]
}
}
}