fieldflow

概述

FieldFlow 是一个基于 Python 的工具，它将 OpenAPI 描述的 REST 端点转换为可选择过滤的工具。它动态生成 Pydantic 模型和 FastAPI 路由，将请求代理到上游 API，同时只返回调用者指定的字段。该项目提供 HTTP 服务器和 MCP 服务器两种模式，后者通过模型上下文协议 (MCP) 暴露 OpenAPI 工具。它支持身份验证方案、JSONPath 类语法的嵌套字段选择，并包含用于减少嘈杂 JSON 输出的 CLI 工具。

试试问 AI

装完之后，这里有 5 个你可以让 AI 做的事：

• get_user_info

  Fetch user information from JSONPlaceholder API with field selection

  试试:Get user 1's name and email试试:Fetch user details for user_id 5 with only the name and email fields
• list_posts

  List posts for a user from JSONPlaceholder API with field selection

  试试:List posts for user 1 with just the title and id试试:Get all posts from user 2 showing only title and creation date
• pokemon_read

  Retrieve Pokemon information from PokeAPI with nested field selection

  试试:Get Pokemon 150's name, types, and attack stat试试:Fetch Bulbasaur's data showing only name and type names
• ability_list

  List Pokemon abilities from PokeAPI with field selection

  试试:List the first 5 abilities with just their names试试:Get abilities showing only result names
• ability_read

  Read a specific Pokemon ability by ID from PokeAPI

  试试:Get ability 65 with name and effect entries试试:Fetch ability details showing only the ability name

说明：Tools are derived from OpenAPI specs and exposed through MCP server. Example tool calls shown in README demonstrate the field selection functionality.

可对比工具

安装

# 安装 MCP 支持
pip install -e '.[mcp]'

# 运行 HTTP 服务器
fieldflow serve-http

# 运行 MCP 服务器（用于 Claude Desktop）
fieldflow-mcp

{
  "mcpServers": {
    "fieldflow": {
      "command": "python",
      "args": ["-m", "fieldflow_mcp.cli"]
    }
  }
}

FAQ

FieldFlow 支持哪些认证方法？

FieldFlow 支持 Bearer 令牌（OAuth 2.0、JWT）、API 密钥、基本身份验证，以及自动处理 OpenAPI 定义的安全方案。

FieldFlow 如何减少响应负载？

FieldFlow 接受请求中的 'fields' 参数，允许选择特定数据字段。它使用 JSONPath 类语法来选择嵌套字段和数组，仅返回请求的数据。

Hacker News 讨论

开发者社区最近的相关讨论。

• Show HN: Transform your REST API into optimized MCP server in one command line▲ 2
  帖子 by ggay · 2025-09-23

fieldflow 对比