Open-source local MCP runtime that checks AI-agent tool calls, gates risky actions on approval, verifies results, and avoids blind retries after timeouts.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add melra -- npx -y @melra/cli~/.codex/config.toml
[mcp_servers.melra]
command = "npx"
args = ["-y", "@melra/cli"]
# codex mcp add melra -- npx -y @melra/cli~/.cursor/mcp.json
{
"mcpServers": {
"melra": {
"command": "npx",
"args": [
"-y",
"@melra/cli"
]
}
}
}.vscode/mcp.json
{
"servers": {
"melra": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@melra/cli"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"melra": {
"command": "npx",
"args": [
"-y",
"@melra/cli"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"melra": {
"command": "npx",
"args": [
"-y",
"@melra/cli"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"melra": {
"command": "npx",
"args": [
"-y",
"@melra/cli"
]
}
}
}