Koma Gate MCP runs on your own machine — the client starts it, so there is no endpoint to ping. 15 installs a week from npm. Last commit 13 Oct 2026.
Classifies prompt injection, jailbreaks, and out-of-scope user input before agents act on it.
We read the source, 5 d ago · rules 995398eba69f
Found in continuous integration, deployment or infrastructure files, or in a neighbouring package of the same monorepo. None of this is installed when you add the server: it describes how the project is built and released. We list it because a leaked key in a build pipeline is still a real problem, but it is not something this server does on your machine.
const result = spawnSync(process.execPath, [live], { stdio: 'inherit' });
const settingsPath = path.join(projectRoot, details.settingsRelativePath);
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
Today is the operative word: we check Koma Gate MCP 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 koma-gate-mcp -- npx -y koma-gate-mcp
{
"mcpServers": {
"koma-gate-mcp": {
"command": "npx",
"args": [
"-y",
"koma-gate-mcp"
]
}
}
}
[mcp_servers.koma-gate-mcp]
command = "npx"
args = ["-y", "koma-gate-mcp"]
{
"mcpServers": {
"koma-gate-mcp": {
"command": "npx",
"args": [
"-y",
"koma-gate-mcp"
]
}
}
}
{
"servers": {
"koma-gate-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"koma-gate-mcp"
]
}
}
}
Optional settings from the registry entry, left out of the block above:
KOMA_PROVIDER (LLM provider: openai, anthropic, google, deepseek, or ollama (default: google)), KOMA_MODEL (Optional model override for the selected provider), GEMINI_API_KEY (Google Gemini API key (used by the default provider)), OPENAI_API_KEY (OpenAI API key when KOMA_PROVIDER=openai), ANTHROPIC_API_KEY (Anthropic API key when KOMA_PROVIDER=anthropic), DEEPSEEK_API_KEY (DeepSeek API key when KOMA_PROVIDER=deepseek).
Detect prompt injection, jailbreaks, and code injection in untrusted text before it reaches an LLM.
Security gateway for AI agents: detects prompt injections, jailbreaks, and common vulnerabilities.
Multi-agent LLM security layer detecting prompt injection and jailbreaks.
Prompt-injection scanner for agents: risk score, categories and spans for text, pages, MCP tools
Detect prompt injection, jailbreak, and social-engineering attacks in LLM agents.
Security gateway for MCP agents: blocks prompt-injection-driven tool calls before they execute.
Real-time MCP & A2A prompt-injection, jailbreak, and tool-schema poisoning firewall for AI agents.
Check an email before an AI agent sends it: send / ask a human / block. Detects prompt injection.
Answers built from our own checks of this server.