KATA Research Automation

Deterministic research automation with live OpenAlex search and reusable workflow programs.

使うべきか

品質と安全性

A
説明の品質
95%
スキーマの完全性
79%
命名の品質
80%
ポイズニングのリスク
100%
権限の一致
90%
プロトコルへの準拠
100%

検出事項(1)

  • LOWTool 'kata_diagnose_web_interop' suggests web access but openWorldHint=falsekata_diagnose_web_interop 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

~1,328トークン数(ツール定義)
~1.4 KB一般的なレスポンスサイズ
注意への影響は中程度(128k コンテキストの 1.04%)

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": {
    "kata": {
      "url": "https://kata-webmcp.vercel.app/api/mcp"
    }
  }
}

リモートエンドポイント

https://kata-webmcp.vercel.app/api/mcpstreamable-http

できること

ツール一覧

ツール(8)

🟢 読み取り専用🟡 書き込み🔴 削除⚪ 不明
🟢kata_search_research(query, limit)

Search live scholarly works through OpenAlex. Returns normalized real research records; never mock data.

入力スキーマ

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "minLength": 1,
      "maxLength": 240
    },
    "limit": {
      "type": "integer",
      "minimum": 1,
      "maximum": 25
    }
  },
  "required": [
    "query"
  ],
  "additionalProperties": false
}
🟢kata_plan_triage(query, limit, works, criteria, actions, ...)

Create a deterministic triage plan from live OpenAlex results or supplied works without mutating a workspace.

入力スキーマ

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "minLength": 1,
      "maxLength": 240
    },
    "limit": {
      "type": "integer",
      "minimum": 1,
      "maximum": 25
    },
    "works": {
      "type": "array",
      "maxItems": 25,
      "items": {
        "type": "object"
      }
    },
    "criteria": {
      "type": "object"
    },
    "actions": {
      "type": "object"
    },
    "maxItems": {
      "type": "integer",
      "minimum": 1,
      "maximum": 25
    }
  },
  "additionalProperties": false
}
🟢kata_diagnose_web_interop(intent, environment)

Classify observed WebMCP, MCP, browser and API constraints and return a conservative compliant integration path without weakening security controls.

入力スキーマ

{
  "type": "object",
  "properties": {
    "intent": {
      "type": "string",
      "enum": [
        "read",
        "act",
        "automate",
        "expose_webmcp",
        "call_api",
        "connect_mcp"
      ]
    },
    "environment": {
      "type": "object",
      "properties": {
        "webMcpApi": {
          "type": "string",
          "enum": [
            "available",
            "unavailable",
            "unknown"
          ]
        },
        "frame": {
          "type": "string",
          "enum": [
            "top",
            "same-origin",
            "cross-origin",
            "unknown"
          ]
        },
        "toolsPermission": {
          "type": "string",
          "enum": [
            "allowed",
            "blocked",
            "unknown"
          ]
        },
        "originExposure": {
          "type": "string",
          "enum": [
            "allowed",
            "blocked",
            "not-required",
            "unknown"
          ]
        },
        "crossOriginRequest": {
          "type": "string",
          "enum": [
            "requested",
            "not-requested",
            "not-required",
            "unknown"
          ]
        },
        "api": {
          "type": "string",
          "enum": [
            "documented",
            "undocumented",
            "none",
            "unknown"
          ]
        },
        "auth": {
          "type": "string",
          "enum": [
            "authenticated",
            "required",
            "none",
            "unknown"
          ]
        },
        "authScope": {
          "type": "string",
          "enum": [
            "browser",
            "all-paths",
            "unknown"
          ]
        },
        "cors": {
          "type": "string",
          "enum": [
            "allowed",
            "blocked",
            "not-applicable",
            "unknown"
          ]
        },
        "cspConnect": {
          "type": "string",
          "enum": [
            "allowed",
            "blocked",
            "not-applicable",
            "unknown"
          ]
        },
        "rateLimit": {
          "type": "string",
          "enum": [
            "ok",
            "limited",
            "unknown"
          ]
        },
        "rateLimitScope": {
          "type": "string",
          "enum": [
            "browser",
            "all-paths",
            "unknown"
          ]
        },
        "botProtection": {
          "type": "string",
          "enum": [
            "clear",
            "challenge",
            "blocked",
            "unknown"
          ]
        },
        "botProtectionScope": {
          "type": "string",
          "enum": [
            "browser",
            "all-paths",
            "unknown"
          ]
        },
        "terms": {
          "type": "string",
          "enum": [
            "allowed",
            "restricted",
            "unknown"
          ]
        },
        "termsScope": {
          "type": "string",
          "enum": [
            "browser",
            "all-paths",
            "unknown"
          ]
        },
        "userAuthorizedBrowserFlow": {
          "type": "boolean"
        },
        "serverSideApiAvailable": {
          "enum": [
            true,
            false,
            "unknown"
          ]
        },
        "mcpEndpoint": {
          "type": "string",
          "enum": [
            "available",
            "protected",
            "legacy-candidate",
            "unavailable",
            "unknown"
          ]
        },
        "mcpModernProtocol": {
          "type": "string",
          "enum": [
            "supported",
            "unsupported",
            "unknown"
          ]
        },
        "mcpAuth": {
          "type": "string",
          "enum": [
            "none",
            "required",
            "unknown"
          ]
        },
        "mcpAuthMetadata": {
          "type": "string",
          "enum": [
            "available",
            "not-required",
            "unverified",
            "unknown"
          ]
        }
      },
      "required": [
        "webMcpApi",
        "frame",
        "toolsPermission",
        "originExposure",
        "api",
        "auth",
        "cors",
        "cspConnect",
        "rateLimit",
        "botProtection",
        "terms",
        "userAuthorizedBrowserFlow"
      ],
      "additionalProperties": false
    }
  },
  "required": [
    "intent",
    "environment"
  ],
  "additionalProperties": false
}
⚪kata_apply_command(workspace, command)

Apply one allowlisted semantic workspace command and return the next workspace snapshot.

入力スキーマ

{
  "type": "object",
  "properties": {
    "workspace": {
      "type": "object"
    },
    "command": {
      "type": "object"
    }
  },
  "required": [
    "workspace",
    "command"
  ],
  "additionalProperties": false
}
🟢kata_compile_workflow(name, demos)

Compile two compatible semantic demonstrations into a deterministic JSON-Schema-constrained program.

入力スキーマ

{
  "type": "object",
  "properties": {
    "name": {
      "type": "string",
      "minLength": 1,
      "maxLength": 128
    },
    "demos": {
      "type": "array",
      "minItems": 2,
      "maxItems": 2,
      "items": {
        "type": "array",
        "minItems": 1,
        "maxItems": 32,
        "items": {
          "type": "object"
        }
      }
    }
  },
  "required": [
    "name",
    "demos"
  ],
  "additionalProperties": false
}
⚪kata_execute_program(workspace, program, input)

Execute a compiled deterministic program transactionally against a supplied workspace snapshot.

入力スキーマ

{
  "type": "object",
  "properties": {
    "workspace": {
      "type": "object"
    },
    "program": {
      "type": "object"
    },
    "input": {
      "type": "object"
    }
  },
  "required": [
    "workspace",
    "program",
    "input"
  ],
  "additionalProperties": false
}
🟢kata_preview_automation(workspace, works, automation)

Preview an automation and return matching works plus an integrity fingerprint. Makes no changes.

入力スキーマ

{
  "type": "object",
  "properties": {
    "workspace": {
      "type": "object"
    },
    "works": {
      "type": "array",
      "maxItems": 25,
      "items": {
        "type": "object"
      }
    },
    "automation": {
      "type": "object"
    }
  },
  "required": [
    "workspace",
    "works",
    "automation"
  ],
  "additionalProperties": false
}
⚪kata_run_automation(workspace, works, automation, previewFingerprint)

Run a deterministic automation only when its preview fingerprint still matches current inputs. Nested tools are depth-bounded.

入力スキーマ

{
  "type": "object",
  "properties": {
    "workspace": {
      "type": "object"
    },
    "works": {
      "type": "array",
      "maxItems": 25,
      "items": {
        "type": "object"
      }
    },
    "automation": {
      "type": "object"
    },
    "previewFingerprint": {
      "type": "string",
      "minLength": 64,
      "maxLength": 64
    }
  },
  "required": [
    "workspace",
    "works",
    "automation",
    "previewFingerprint"
  ],
  "additionalProperties": false
}

コミュニティ

このサーバーを評価する

エビデンス

最近の観測

検証済みバージョンは記録されていませんツール 8 件
検証済みバージョンは記録されていませんツール 8 件