Customer Info Validator
Validates customer information completeness by checking caller-provided data against required fields, returning present and missing fields.
README
Customer Info Validator
Customer Info Validator is a Gate 2 Cloudflare Workers + TypeScript MCP app for checking whether caller-provided customer information contains all caller-specified required fields.
App Identity
- App Name: Customer Info Validator
- App Slug: customer-info-validator
- Version: 1.0.0
- App Type: Validator
- Hub: Sales / Customer Operations Hub
- GitHub Repo: customer-info-validator
- Cloudflare Worker Name: customer-info-validator
- Wrangler name: customer-info-validator
- Website URL: https://customer-info-validator.sidcraigau.workers.dev
- MCP URL: https://customer-info-validator.sidcraigau.workers.dev/mcp
- Support Email: sidcraigau@gmail.com
Use Case
The only supported use case is validating customer information completeness. The tool accepts a customer information object and a caller-provided required field list, then returns checked fields, present fields, missing fields, completeness status, source label, and structured errors.
MCP Tool
- Tool Name:
validate_customer_info_completeness - Tool Title: Validate Customer Information Completeness
Input
customer_info: object, required, allows arbitrary caller-provided customer fields.required_fields: array of strings, required, runtime validates empty arrays and blank field names.source_label: string, optional.request_intent: string, optional, used only to reject out-of-scope requests.
Output
Every tool response returns the same top-level structure:
status:successorerrorcomplete: booleanrequired_fields: string arraychecked_fields: string arraypresent_fields: string arraymissing_fields: string arraysource_label: stringerrors: structured error array
Error Contract
Supported error codes:
missing_required_inputinvalid_input_typeempty_inputout_of_scopeinternal_error
Each error object includes code, message, and field.
Out of Scope
The app does not infer or fabricate missing information, score customers, predict conversion, create sales strategies, contact customers, write to CRM systems, create follow-up tasks, generate or send quotations, approve, publish, submit, purchase, transfer, delete, call external APIs, or modify external systems.
Routes
GET /: minimal public HTML page.GET /health: returns{ "status": "ok" }.POST /mcp: JSON-RPC MCP endpoint supportinginitialize,tools/list, andtools/call.
The app does not use /sse, /, or /{app-slug}/mcp as an MCP endpoint.
Local Commands
npm install
npm run typecheck
npm run dev
npm run test:mcp:local
npm run test:mcp:local starts the local Worker, waits for health, exercises the HTTP and MCP routes, validates tool annotations, runs positive and error cases, checks unknown tool behavior, validates structuredContent against outputSchema, and exits nonzero on failure.
Safety Architecture
Current architecture is read-only, stateless, and local to caller-provided input. It has no login, OAuth, database, external API client, CRM write path, message sending, form submission, approval action, purchase action, transfer action, or external system mutation.
Schema / Runtime Boundary
The JSON Schema defines field names, primitive types, arrays, objects, and top-level additionalProperties: false. Runtime type validation still checks every argument and field type. Business validation handles empty customer_info, empty required_fields, blank field names, field trimming, de-duplication, valid content checks, and out-of-scope request_intent detection so business errors are returned through the structured Error Contract.
Processing order:
- arguments
- runtime type validation
- business validation
- deterministic completeness logic
- unified response builder
Deployment Sync Rule
Any future production code change must follow:
Code Patch -> Build -> Deploy -> Production Verification
Patch-after-deploy drift is not allowed. Formal deployment and production verification are not part of Gate 2 and must be performed only in the appropriate later gate.
Gate Status
This repository is currently scoped to Gate 2 Build + Connect. It has not performed Gate 2.5 Developer Mode real listing tests, formal production deployment, formal submission, or final audit approval.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。