Qorami runs on your own machine — the client starts it, so there is no endpoint to ping. 38 installs a week from npm. Last commit 3 Jul 2026.
Check an email before an AI agent sends it: send / ask a human / block. Detects prompt injection.
Today is the operative word: we check Qorami 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 qorami -e 'QORAMI_API_KEY=<QORAMI_API_KEY>' -- npx -y qorami-mcp
{
"mcpServers": {
"qorami": {
"command": "npx",
"args": [
"-y",
"qorami-mcp"
],
"env": {
"QORAMI_API_KEY": "<QORAMI_API_KEY>"
}
}
}
}
[mcp_servers.qorami]
command = "npx"
args = ["-y", "qorami-mcp"]
env = { QORAMI_API_KEY = "<QORAMI_API_KEY>" }
{
"mcpServers": {
"qorami": {
"command": "npx",
"args": [
"-y",
"qorami-mcp"
],
"env": {
"QORAMI_API_KEY": "<QORAMI_API_KEY>"
}
}
}
}
{
"servers": {
"qorami": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"qorami-mcp"
],
"env": {
"QORAMI_API_KEY": "<QORAMI_API_KEY>"
}
}
}
}
Put your own value in place of
<QORAMI_API_KEY> (Your Qorami workspace API key (get one at https://qorami.fr/dashboard/)).
The author’s registry entry marks it as required.
Check an email before your AI agent sends it: allow, review or block with reasons.
A real inbox for AI agents: send, receive and thread email, behind a prompt-injection firewall.
A real inbox for AI agents: send, receive and thread email, behind a prompt-injection firewall.
Check a vendor (name, domain, email) for fraud signals and sanctions before an agent pays it.
A real email inbox for AI agents: human-approved sends, untrusted-by-default inbound.
Email safety for AI agents: scan inbound for prompt-injection/phishing + outbound for secret leaks.
Email verification for AI agents: check addresses and lists, priced before credits are spent.
Email inboxes for AI agents: send, reply, read verification codes, human-approved drafts.
Answers built from our own checks of this server.