MCP Operation Risk
Operation-level risk guidance for consolidated MCP tools, including schema drift and retry signals.
使うべきか
品質と安全性
検出事項（1）
- LOWTool 'describe_operation_risk_service' name length outside 3-30 rangedescribe_operation_risk_service 内
ツール定義とプロトコルへの準拠に関する自動分析に基づいています。
コンテキストコスト
これは、サーバーのツールがモデルのコンテキストに読み込まれるたびに消費されるおおよそのトークン数です。数が多いほど、ほかのタスクに使える注意が減ります。
インストール
ワンクリックインストール
これを `claude_desktop_config.json` ファイルに追加してください:
{
"mcpServers": {
"mcp-operation-risk": {
"url": "https://87-106-119-237.sslip.io/mcp-operation-risk/mcp"
}
}
}リモートエンドポイント
https://87-106-119-237.sslip.io/mcp-operation-risk/mcpstreamable-httpできること
ツール一覧
ツール（2）
🟢describe_operation_risk_service
Check whether the actual operation inside an MCP tool call is safe before execution. Use this when one MCP tool exposes multiple operations with different side effects, so whole-tool permissions are too coarse. Provide the native MCP server identity, the tool's tools/list metadata and the intended arguments. For supported consolidated tools, distinguish read, create/add, update/mutate and delete/destructive operations. Return ALLOW, DENY, or REQUIRE_APPROVAL. Unknown, unsupported, ambiguous, or schema-changed calls fail closed with REQUIRE_APPROVAL. The evaluator never executes the source operation.
入力スキーマ
{
"type": "object",
"properties": {},
"title": "describe_operation_risk_serviceArguments"
}出力スキーマ
{
"type": "object",
"additionalProperties": true,
"title": "describe_operation_risk_serviceDictOutput"
}🟢evaluate_operation_risk(source, tool, arguments)
Check whether the actual operation inside an MCP tool call is safe before execution. Use this when one MCP tool exposes multiple operations with different side effects, so whole-tool permissions are too coarse. Provide the native MCP server identity, the tool's tools/list metadata and the intended arguments. For supported consolidated tools, distinguish read, create/add, update/mutate and delete/destructive operations. Return ALLOW, DENY, or REQUIRE_APPROVAL. Unknown, unsupported, ambiguous, or schema-changed calls fail closed with REQUIRE_APPROVAL. The evaluator never executes the source operation.
入力スキーマ
{
"type": "object",
"properties": {
"source": {
"additionalProperties": false,
"properties": {
"server": {
"additionalProperties": false,
"properties": {
"name": {
"description": "Name reported by the MCP server during initialize; used as a supported-manifest lookup hint.",
"title": "Name",
"type": "string"
},
"version": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Server version reported during initialize, or null when unavailable.",
"title": "Version"
}
},
"required": [
"name",
"version"
],
"title": "NativeServerIdentity",
"type": "object",
"description": "MCP server identity from the connection's initialize response."
},
"protocol_version": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Negotiated MCP protocol version, or null when unavailable.",
"title": "Protocol Version"
}
},
"required": [
"server",
"protocol_version"
],
"title": "NativeSource",
"type": "object",
"description": "MCP server and negotiated protocol identity for the source tool."
},
"tool": {
"additionalProperties": false,
"properties": {
"name": {
"description": "Exact tool name from the MCP server's tools/list response.",
"title": "Name",
"type": "string"
},
"title": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"description": "Tool title from tools/list, when provided by the server.",
"title": "Title"
},
"description": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"description": "Tool description from tools/list, or null if the server omitted it.",
"title": "Description"
},
"inputSchema": {
"additionalProperties": true,
"description": "Complete inputSchema from tools/list; used to validate the intended arguments and match a supported tool manifest.",
"title": "Inputschema",
"type": "object"
},
"annotations": {
"additionalProperties": true,
"description": "Tool annotations from tools/list, such as read-only or destructive hints; use an empty object when absent.",
"title": "Annotations",
"type": "object"
},
"execution": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Tool execution metadata from tools/list, when present.",
"title": "Execution"
},
"outputSchema": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Tool output schema from tools/list, when present.",
"title": "Outputschema"
},
"icons": {
"anyOf": [
{
"items": {
"additionalProperties": true,
"type": "object"
},
"type": "array"
},
{
"type": "null"
}
],
"default": null,
"description": "Tool icons from tools/list, when present.",
"title": "Icons"
},
"_meta": {
"anyOf": [
{
"additionalProperties": true,
"type": "object"
},
{
"type": "null"
}
],
"default": null,
"description": "Relevant tool metadata from tools/list, when present.",
"title": "Meta"
}
},
"required": [
"name",
"description",
"inputSchema"
],
"title": "NativeToolRecord",
"type": "object",
"description": "The complete relevant Tool record returned by the source server's tools/list."
},
"arguments": {
"additionalProperties": true,
"description": "Arguments intended for the source MCP tools/call request being evaluated. The evaluator does not execute that call.",
"title": "Arguments",
"type": "object"
}
},
"required": [
"source",
"tool",
"arguments"
],
"additionalProperties": false,
"title": "NativeMcpEvaluationRequest"
}コミュニティ
エビデンス