From sentinel-ai
Runs locally via Python. Sentinel AI safety scanner: detects prompt injection, PII, harmful/toxic content, hallucinations in text; audits tool calls for dangerous operations.
npx claudepluginhub maxwellcalkin/sentinel-ai --plugin sentinel-aiAdd to your .mcp.json:
{
"mcpServers": {
"sentinel-ai": {
"command": "python",
"args": [
"-m",
"sentinel.mcp_server"
]
}
}
}python