SendGrid Secure runs on your own machine — the client starts it, so there is no endpoint to ping. 159 installs a week from pypi. Last commit 2 Sep 2026.
Security-first SendGrid MCP server: dry-run default, recipient allowlists, rate caps, audit log.
Today is the operative word: we check SendGrid Secure every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add sendgrid-mcp-secure -- uvx sendgrid-mcp-secure
{
"mcpServers": {
"sendgrid-mcp-secure": {
"args": [
"sendgrid-mcp-secure"
],
"command": "uvx"
}
}
}
[mcp_servers.sendgrid-mcp-secure]
command = "uvx"
args = ["sendgrid-mcp-secure"]
{
"mcpServers": {
"sendgrid-mcp-secure": {
"args": [
"sendgrid-mcp-secure"
],
"command": "uvx"
}
}
}
{
"mcpServers": {
"sendgrid-mcp-secure": {
"args": [
"sendgrid-mcp-secure"
],
"command": "uvx"
}
}
}
This one needs environment variables set before it will start:
SENDGRID_API_KEY (SendGrid API key (use a restricted key with Mail Send scope only)), SENDGRID_MCP_MODE (dry-run (default) or live), SENDGRID_MCP_RECIPIENT_ALLOWLIST (Comma-separated recipient addresses/domains allowed to receive mail), SENDGRID_MCP_WRITES_PER_HOUR (Hourly cap on write operations (sends)), SENDGRID_MCP_MAX_RECIPIENTS (Max recipients per send), SENDGRID_MCP_AUDIT_LOG (Path to the append-only audit log), SENDGRID_MCP_ALLOW_BCC (BCC stays disabled unless explicitly enabled).
The author declared them in the registry entry; get the values from the project itself.
Security gate for agent-driven WhatsApp: allowlist, secret scan, rate limit, audit log.
Safe-by-default UniFi MCP: Network + Protect + Access, multi-site, dry-run, audit log.
Local-first secret scanning, rotation, vault, and audit-log tools for AI agents.
MCP server for MikroTik RouterOS: typed tools, dry-run, RBAC, audit logs, and rollback.
MCP policy proxy: spend caps, approvals for destructive tools, kill switch, dry-run, audit log.
Secure SQL proxy for AI agents — NL→SQL, AST safety, per-agent RLS, audit log.
Local-first MCP server for data-quality discovery, evidence, drift, and controlled repair.
Security scanner and runtime proxy for MCP servers.
Answers built from our own checks of this server.