gotoHuman MCP
OfficialgotoHuman MCP 服务器
gotoHuman 可以轻松地将人工批准添加到 AI 代理和代理工作流程中。
具有可定制批准 UI 的完全管理的异步人机循环工作流程。
享受内置身份验证、webhook、通知、团队功能和不断发展的训练数据集。
使用我们的 MCP 服务器通过 MCP 从您的 AI 工作流请求人工批准,或将其添加到您的 IDE 以帮助集成。
安装
npx @gotohuman/mcp-server与 Cursor / Claude / Windsurf 一起使用
{
"mcpServers": {
"gotoHuman": {
"command": "npx",
"args": ["-y", "@gotohuman/mcp-server"],
"env": {
"GOTOHUMAN_API_KEY": "your-api-key"
}
}
}
}获取您的 API 密钥并在app.gotohuman.com设置审批步骤
Related MCP server: final-approval
演示
这是左侧的光标,但它可能是对批准 webhook 做出反应的后台代理。
https://github.com/user-attachments/assets/380a4223-ea77-4e24-90a5-52669b77f56f
工具
list-forms
列出所有可用的审查表格。
返回您帐户中所有可用表单的列表,包括有关添加字段的高级信息
get-form-schema
获取请求对给定表单进行人工审核时使用的模式。
参数
formId:用于获取架构的表单 ID
返回模式,考虑包含的字段及其配置
request-human-review-with-form
申请人工审核。邮件将出现在您的 gotoHuman 收件箱中。
参数
formId:评论的表单 IDfieldData:表单字段的内容(AI 输出以进行审查、上下文……)和配置。
需要使用get-form-schema获取此模式metadata:表单提交后,将包含在 webhook 响应中的可选附加数据assignToUsers评论的用户电子邮件的可选列表
返回gotoHuman 中评论的链接
发展
# Install dependencies
npm install
# Build the server
npm run build
# For testing: Run the MCP inspector
npm run inspector在 MCP 客户端本地运行(例如 Cursor / Claude / Windsurf)
{
"mcpServers": {
"gotoHuman": {
"command": "node",
"args": ["/<absolute-path>/build/index.js"],
"env": {
"GOTOHUMAN_API_KEY": "your-api-key"
}
}
}
}[!NOTE] 对于 Windows,
args路径需要为C:\\<absolute-path>\\build\\index.js
Available Tools
3 toolsget-form-schemaA
Get the schema to use for the 'fields' property when requesting a human review with a form.
Inputs: formId (required).
list-formsA
List all available review forms. NOTE: You need to fetch the schema for the form fields first using the get-form-schema tool.
Inputs: no named parameters.
request-human-review-with-formA
Request a human review with a form. NOTE: If you don't have a form ID yet, list all available forms using the list-forms tool first. To know what to pass for fieldData, you need to fetch the schema for the form fields using the get-form-schema tool.
Inputs: formId (required), fieldData (required), metadata, assignToUsers.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 3 tools
Each tool has a clearly distinct purpose: get-form-schema retrieves schema information, list-forms enumerates available forms, and request-human-review-with-form submits a review request. There is no overlap in functionality, and the descriptions explicitly guide agents on when to use each tool, preventing misselection.
The tool names follow a consistent verb-noun pattern with hyphens (e.g., get-form-schema, list-forms, request-human-review-with-form), making them readable and predictable. The minor deviation is that request-human-review-with-form includes a prepositional phrase, but this does not significantly impact consistency.
With 3 tools, this server is well-scoped for its purpose of handling human reviews with forms. Each tool earns its place by covering essential steps: schema retrieval, form listing, and review submission, without being overly sparse or bloated.
The tool set provides complete coverage for the domain of requesting human reviews with forms. It includes all necessary CRUD-like operations: retrieving schema (read), listing forms (read), and submitting reviews (create), with no obvious gaps that would cause agent failures in this workflow.
Maintenance
Related MCP Connectors
Give any AI agent a way to ask a person — for approval, a decision, an answer or a review.
Set up human approval for your AI agents' tool calls.
Human approvals, notifications, inbound webhooks, wake-ups for headless agents. Free trial: /try
Human-in-the-loop API for AI agents. CAPTCHA, OTP, KYC, and approvals by real humans.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceHuman-in-the-Loop authorization gateway for AI Agents. Securely pause MCP workflows and route high-risk actions to human approvers via Slack or Email.56 npm1MIT
- AlicenseAqualityDmaintenanceHuman-in-the-loop approval gate for AI agents. Your agent calls submit_approval before any irreversible action; a human reviews on a branded page; a signed webhook fires back with the decision.11MIT
- AlicenseNot gradedqualityDmaintenanceProvides a human approval gate for AI agents, enabling interactive inline cards for approving, editing, or rejecting actions before they are executed.MIT
- AlicenseNot gradedqualityDmaintenancePauses AI agent execution and routes approval requests to humans via Slack or email, with cryptographically signed proof of the human's decision.MIT