mcp-vtenext
mcp-vtenext
用于 VTENext CRM 的 MCP 服务器 — 将 WebService API 作为工具暴露给 Claude 和其他兼容 MCP 的客户端。
要求
Node.js 18+
一个正在运行的 VTENext 实例(自托管或 Docker — 参见 ../docker)
Related MCP server: nimble-crm-mcp
设置
cd mcp/vtenext/server
npm install
cp .env.example .env编辑 .env:
VTENEXT_URL=http://your-vtenext-instance
VTENEXT_USERNAME=admin
VTENEXT_ACCESS_KEY=your_access_key
READ_ONLY=false访问密钥位于 VTENext 的 管理 (Admin) → 用户 (Users) → [用户] → 访问密钥 (Access Key) 中。
只读模式
设置 READ_ONLY=true 以防止对 VTENext 进行任何写入操作。启用后,工具 create_opportunita、update_opportunita 和 add_nota_opportunita 将返回错误,而不是写入数据。
当服务器被 AI 机器人或自动化代理使用,且仅需读取 CRM 数据时,此功能非常有用。若要与全权限实例同时运行只读实例,请通过 MCP 配置传递该变量:
{
"mcpServers": {
"vtenext-bot": {
"type": "stdio",
"command": "node",
"args": ["/absolute/path/to/mcp/vtenext/server/index.js"],
"env": {
"VTENEXT_URL": "http://your-vtenext-instance",
"VTENEXT_USERNAME": "admin",
"VTENEXT_ACCESS_KEY": "your_access_key",
"READ_ONLY": "true"
}
}
}
}Claude Code 集成
添加到项目根目录的 .mcp.json 中:
{
"mcpServers": {
"vtenext": {
"type": "stdio",
"command": "node",
"args": ["/absolute/path/to/mcp/vtenext/server/index.js"]
}
}
}工具
Opportunità (商机)
工具 | 描述 |
| 列出商机,支持可选过滤器(状态、搜索、限制) |
| 通过 ID 获取商机的详细信息 |
| 按名称搜索商机 |
| 创建新商机 (写入操作 — 在只读模式下被阻止) |
| 更新现有商机的状态、金额或备注 (写入操作 — 在只读模式下被阻止) |
Contatti (联系人)
工具 | 描述 |
| 按姓名、电子邮件或公司搜索联系人 |
Attività e note (活动与备注)
工具 | 描述 |
| 向商机添加评论/备注 (写入操作 — 在只读模式下被阻止) |
| 列出与商机关联的活动 |
Utilità (实用工具)
工具 | 描述 |
| 显示任何 VTENext 模块的可用字段 |
| 运行原始 VTQL SELECT 查询 |
身份验证
VTENext 使用 vtiger WebService 协议:
GET /webservice.php?operation=getchallenge→ tokenMD5(token + accessKey) → 哈希密钥
POST /webservice.php使用operation=login(表单编码) → sessionName
会话缓存时间为 4 分钟(令牌有效期为 5 分钟)。
测试
# Unit tests (no VTENext required)
npm test
# Integration tests (requires live VTENext at VTENEXT_URL)
npm run test:integration许可证
MIT
Available Tools
10 toolsadd_nota_opportunitaC
Aggiunge una nota/commento a un'opportunità
Inputs: opportunita_id (required), testo (required).
create_opportunitaC
Crea una nuova opportunità in VTENext
Inputs: nome (required), stato, importo, data_chiusura, +1 more.
describe_moduloC
Mostra i campi disponibili per un modulo VTENext
Inputs: modulo (required).
get_opportunitaC
Recupera i dettagli completi di un'opportunità tramite ID
Inputs: id (required).
list_attivita_opportunitaC
Elenca le attività associate a un'opportunità
Inputs: opportunita_id (required), limit.
list_opportunitaC
Elenca le opportunità con filtri opzionali
Inputs: limit, stato, search.
query_rawC
Esegue una query VTQL grezza su VTENext (solo lettura)
Inputs: query (required).
search_contattiC
Cerca contatti per nome, email o azienda
Inputs: q (required), limit.
search_opportunitaC
Cerca opportunità per nome
Inputs: q (required), limit.
update_opportunitaC
Aggiorna i dati di un'opportunità esistente
Inputs: id (required), stato, importo, descrizione.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 10 tools
Each tool has a clearly distinct purpose with no ambiguity. For example, list_opportunita retrieves multiple records with filters, get_opportunita fetches a single record by ID, and search_opportunita finds records by name. The tools target different resources (opportunità, contatti, modulo) and actions (create, update, list, search, describe, query).
All tool names follow a consistent verb_noun pattern in snake_case. Verbs like add, create, describe, get, list, query, search, and update are used predictably with corresponding nouns (e.g., add_nota_opportunita, create_opportunita, describe_modulo). There are no deviations in naming conventions.
With 10 tools, the count is well-scoped for managing opportunities and related data in VTENext. Each tool serves a clear purpose, such as CRUD operations for opportunities, searching contacts, listing activities, and raw queries. No tool feels redundant or unnecessary for the server's domain.
The tool set provides complete coverage for the VTENext opportunity management domain. It includes full CRUD (create, get, update, list/search) for opportunities, plus related operations like adding notes, listing activities, searching contacts, describing modules, and raw queries. There are no obvious gaps, and agents can handle typical workflows without dead ends.
Maintenance
Related MCP Connectors
API-first CRM for LLMs - contacts, companies, deals and activities over a native MCP server.
Vtiger MCP server for ChatGPT, Claude and any AI agent: list contacts, convert contact to deal, list deals, update deal stage, list cases, create case, update case and create contact in your own Vtiger account. Cotlet works in the account through your Cloud Browser, a real Chrome where you sign in to Vtiger yourself; Cotlet does not store your password and you need no Vtiger API key. About the official Vtiger API: Yes, Vtiger has a REST API, but its limits table lists 0 requests for One Pilot and 24,000 a day (30 a minute) for One Growth, and Vtiger has no official MCP server (checked 22 Oct 2026). Setup and FAQ: https://cotlet.dev/integrations/vtiger-api-mcp-server
MCP server unifying ERPs, CRMs, APIs and knowledge base for Claude, ChatGPT and Gemini.
MCP server for Getfly CRM — customers, opportunities, orders, tickets and tasks.
Related MCP Servers
- AlicenseNot gradedqualityNot gradedmaintenanceEnables users to manage CRM leads, opportunities, projects, tasks, and sales records within ERPNext directly through MCP-compatible clients. It supports global searches, timesheet entries, and access to dashboard KPIs using natural language.-
- AlicenseNot gradedqualityDmaintenanceMCP server for the Nimble CRM API, enabling management of contacts, deals, notes, tasks, and tags through natural language.MIT
- AlicenseNot gradedqualityCmaintenanceMCP server for vtiger CRM that enables AI agents to log timesheets and create/update records in any module via natural language.33 npm1MIT
- AlicenseAqualityDmaintenanceMCP server that connects to a sales CRM database, enabling AI agents to search prospects, view pipeline status, manage follow-ups, and update records using natural language commands.5MIT