ai-safety-guard
Ensure AI agent safety by analyzing and providing security guidelines for operations involving emails, databases, files, and more. Classifies actions by sensitivity and operation type.
Instructions
AI Safety Guard - MCP Caution Instructions for AI Agents
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| mcp_type | No | Type of MCP the AI Agent is about to call | general |
| operation_type | No | Type of operation being requested | read |
| sensitivity_level | No | Sensitivity level of the data/operation | internal |
Input Schema (JSON Schema)
{
"$schema": "http://json-schema.org/draft-07/schema#",
"additionalProperties": false,
"properties": {
"mcp_type": {
"default": "general",
"description": "Type of MCP the AI Agent is about to call",
"enum": [
"email",
"slack",
"database",
"file",
"web",
"general"
],
"type": "string"
},
"operation_type": {
"default": "read",
"description": "Type of operation being requested",
"enum": [
"read",
"write",
"execute",
"delete",
"send",
"query"
],
"type": "string"
},
"sensitivity_level": {
"default": "internal",
"description": "Sensitivity level of the data/operation",
"enum": [
"public",
"internal",
"confidential",
"restricted"
],
"type": "string"
}
},
"type": "object"
}