Pathwize

Run data annotation and evaluation tasks with vetted domain experts: create, publish, get results.

使うべきか

品質と安全性

A
説明の品質
100%
スキーマの完全性
71%
命名の品質
100%
ポイズニングのリスク
100%
権限の一致
100%
プロトコルへの準拠
100%

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

~1,793トークン数(ツール定義)
~1.7 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 1.40%)

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": {
    "pathwize": {
      "url": "https://www.gopathwize.com/api/mcp"
    }
  }
}

リモートエンドポイント

https://www.gopathwize.com/api/mcpstreamable-http

できること

ツール一覧

ツール(10)

🟢 読み取り専用🟡 書き込み🔴 削除⚪ 不明
🟢list_pools

Pools a task can run on: active curated pools plus your own, with member counts.

入力スキーマ

{
  "type": "object",
  "properties": {}
}
🟢list_tasks

Your AI tasks, newest first, with status and item counts.

入力スキーマ

{
  "type": "object",
  "properties": {}
}
🟢get_task(taskId)

One task with settings, classes, examples, qualification, reference item count and item counts by status.

入力スキーマ

{
  "type": "object",
  "properties": {
    "taskId": {
      "type": "string",
      "format": "uuid"
    }
  },
  "required": [
    "taskId"
  ]
}
🟡create_task(title, type, instructions, classes, annotation, ...)

Create a labeling or evaluation task as a draft; nobody is notified until publish_task. Typical order: list_pools, create_task with instructions, classes, rate, pool_ids and one example image per class, add_items, then publish_task after the user confirms.

入力スキーマ

{
  "type": "object",
  "properties": {
    "title": {
      "type": "string",
      "minLength": 1,
      "maxLength": 200
    },
    "type": {
      "type": "string",
      "enum": [
        "data_labeling",
        "preference",
        "evaluation",
        "red_team",
        "domain"
      ],
      "description": "Default data_labeling"
    },
    "instructions": {
      "type": "string",
      "description": "Instructions for the experts in Markdown. A table whose first column names the classes becomes the per-class briefing."
    },
    "classes": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "maxItems": 50,
      "description": "Class names experts choose from"
    },
    "annotation": {
      "type": "string",
      "enum": [
        "classification",
        "bounding_box",
        "segmentation",
        "attributes"
      ],
      "description": "Default classification"
    },
    "rate": {
      "type": "number",
      "minimum": 5,
      "maximum": 500,
      "description": "Hourly pay for the expert in EUR"
    },
    "pool_ids": {
      "type": "array",
      "items": {
        "type": "string",
        "format": "uuid"
      },
      "description": "Pools the task runs on (see list_pools)"
    },
    "members_only": {
      "type": "boolean",
      "description": "Only members of these pools see the task, even on public pools"
    },
    "domain": {
      "type": "string"
    },
    "expertise": {
      "type": "string",
      "description": "What experts need to know, one line"
    },
    "multi_select": {
      "type": "boolean"
    },
    "allow_note": {
      "type": "boolean"
    },
    "allow_cant_tell": {
      "type": "boolean",
      "description": "Let experts skip an item without a class (default true)"
    },
    "seconds_per_item": {
      "type": "integer",
      "minimum": 1,
      "maximum": 3600
    },
    "language": {
      "type": "string",
      "enum": [
        "en",
        "de"
      ],
      "description": "Language of the instructions"
    },
    "deadline": {
      "type": "string",
      "pattern": "^\\d{4}-\\d{2}-\\d{2}$"
    },
    "qualification": {
      "type": "object",
      "properties": {
        "enabled": {
          "type": "boolean"
        },
        "sample": {
          "type": "integer",
          "minimum": 5,
          "maximum": 1000,
          "description": "Test items per candidate (default 50)"
        },
        "scoring": {
          "type": "string",
          "enum": [
            "auto",
            "customer"
          ],
          "description": "auto: scored against reference labels; customer: you approve each candidate"
        },
        "threshold": {
          "type": "integer",
          "minimum": 50,
          "maximum": 100,
          "description": "Pass mark in percent"
        },
        "days_to_complete": {
          "type": "integer",
          "minimum": 1,
          "maximum": 30
        }
      }
    },
    "examples": {
      "type": "array",
      "maxItems": 40,
      "items": {
        "type": "object",
        "properties": {
          "url": {
            "type": "string",
            "format": "uri"
          },
          "label": {
            "type": "string"
          }
        },
        "required": [
          "url",
          "label"
        ]
      },
      "description": "Example images by public URL with the class they show. Labeling tasks need one per class before they can open."
    }
  },
  "required": [
    "title"
  ]
}
🟡update_task(taskId, title, instructions, classes, annotation, ...)

Change a task's settings; only the fields you pass change, examples are added. Does not publish.

入力スキーマ

{
  "type": "object",
  "properties": {
    "taskId": {
      "type": "string",
      "format": "uuid"
    },
    "title": {
      "type": "string"
    },
    "instructions": {
      "type": "string",
      "description": "Instructions for the experts in Markdown. A table whose first column names the classes becomes the per-class briefing."
    },
    "classes": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "maxItems": 50,
      "description": "Class names experts choose from"
    },
    "annotation": {
      "type": "string",
      "enum": [
        "classification",
        "bounding_box",
        "segmentation",
        "attributes"
      ],
      "description": "Default classification"
    },
    "rate": {
      "type": "number",
      "minimum": 5,
      "maximum": 500,
      "description": "Hourly pay for the expert in EUR"
    },
    "pool_ids": {
      "type": "array",
      "items": {
        "type": "string",
        "format": "uuid"
      },
      "description": "Pools the task runs on (see list_pools)"
    },
    "members_only": {
      "type": "boolean",
      "description": "Only members of these pools see the task, even on public pools"
    },
    "domain": {
      "type": "string"
    },
    "expertise": {
      "type": "string",
      "description": "What experts need to know, one line"
    },
    "multi_select": {
      "type": "boolean"
    },
    "allow_note": {
      "type": "boolean"
    },
    "allow_cant_tell": {
      "type": "boolean",
      "description": "Let experts skip an item without a class (default true)"
    },
    "seconds_per_item": {
      "type": "integer",
      "minimum": 1,
      "maximum": 3600
    },
    "language": {
      "type": "string",
      "enum": [
        "en",
        "de"
      ],
      "description": "Language of the instructions"
    },
    "deadline": {
      "type": "string",
      "pattern": "^\\d{4}-\\d{2}-\\d{2}$"
    },
    "qualification": {
      "type": "object",
      "properties": {
        "enabled": {
          "type": "boolean"
        },
        "sample": {
          "type": "integer",
          "minimum": 5,
          "maximum": 1000,
          "description": "Test items per candidate (default 50)"
        },
        "scoring": {
          "type": "string",
          "enum": [
            "auto",
            "customer"
          ],
          "description": "auto: scored against reference labels; customer: you approve each candidate"
        },
        "threshold": {
          "type": "integer",
          "minimum": 50,
          "maximum": 100,
          "description": "Pass mark in percent"
        },
        "days_to_complete": {
          "type": "integer",
          "minimum": 1,
          "maximum": 30
        }
      }
    },
    "examples": {
      "type": "array",
      "maxItems": 40,
      "items": {
        "type": "object",
        "properties": {
          "url": {
            "type": "string",
            "format": "uri"
          },
          "label": {
            "type": "string"
          }
        },
        "required": [
          "url",
          "label"
        ]
      },
      "description": "Example images by public URL with the class they show. Labeling tasks need one per class before they can open."
    }
  },
  "required": [
    "taskId"
  ]
}
🟡add_items(taskId, reference, items)

Add items by public URL; Pathwize copies the files into its own storage. reference: true adds them to the qualification set (optionally with gold_label for automatic scoring). Idempotent per id, up to 1000 per call.

入力スキーマ

{
  "type": "object",
  "properties": {
    "taskId": {
      "type": "string",
      "format": "uuid"
    },
    "reference": {
      "type": "boolean"
    },
    "items": {
      "type": "array",
      "minItems": 1,
      "maxItems": 1000,
      "items": {
        "type": "object",
        "properties": {
          "url": {
            "type": "string",
            "format": "uri"
          },
          "id": {
            "type": "string",
            "description": "Defaults to the file name in the URL"
          },
          "mime": {
            "type": "string",
            "description": "Defaults from the file extension"
          },
          "gold_label": {
            "type": "string"
          }
        },
        "required": [
          "url"
        ]
      }
    }
  },
  "required": [
    "taskId",
    "items"
  ]
}
🟡publish_task(taskId, confirm)

Open a task for experts. This emails every expert in the task's pools once and cannot be undone, so call it only after the user explicitly confirmed, with confirm: true. Fails with a list of anything missing.

入力スキーマ

{
  "type": "object",
  "properties": {
    "taskId": {
      "type": "string",
      "format": "uuid"
    },
    "confirm": {
      "type": "boolean",
      "const": true
    }
  },
  "required": [
    "taskId",
    "confirm"
  ]
}
⚪close_task(taskId, status)

closed ends the task, draft pauses it. Experts are not emailed.

入力スキーマ

{
  "type": "object",
  "properties": {
    "taskId": {
      "type": "string",
      "format": "uuid"
    },
    "status": {
      "type": "string",
      "enum": [
        "closed",
        "draft"
      ]
    }
  },
  "required": [
    "taskId"
  ]
}
🟢get_results(taskId, limit, starting_after, since)

Finished items with final label, agreement, notes and flags. Page with starting_after (the last item id).

入力スキーマ

{
  "type": "object",
  "properties": {
    "taskId": {
      "type": "string",
      "format": "uuid"
    },
    "limit": {
      "type": "integer",
      "minimum": 1,
      "maximum": 1000
    },
    "starting_after": {
      "type": "string"
    },
    "since": {
      "type": "string"
    }
  },
  "required": [
    "taskId"
  ]
}
⚪open_dashboard(taskId)

One-time link that opens the Pathwize business dashboard signed in as this business (optionally one task). Single use, one hour; give it only to the user.

入力スキーマ

{
  "type": "object",
  "properties": {
    "taskId": {
      "type": "string",
      "format": "uuid"
    }
  }
}

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 10 件