servicenow-platform-mcp
Enables AI agents to interact with ServiceNow through MCP, providing schema inspection, record CRUD, attachments, audit, Flow Designer, and platform investigations with policy-guarded access.
README
servicenow-platform-mcp
A Python 3.12+ async MCP server that gives AI agents structured, policy-guarded access to ServiceNow. It exposes schema inspection, record CRUD, attachment operations, audit trail analysis, Flow Designer inspection, and platform investigations - all over the Model Context Protocol's stdio transport.
What you get
The server presents a unified tool surface to any MCP-compatible client. Capabilities include:
- Query and inspect - encoded-query search, aggregation, field-level schema discovery, choice-label resolution.
- Read and write records - including script-bearing artifacts (Business Rules, Script Includes, UI Macros) with file-based script injection and preview-then-apply confirmation.
- Attachments - list, download, upload, and delete with size-capped base64 transfer.
- Audit - field-level audit verdict resolution, batch checks, and masked history retrieval.
- Flow Designer - inspect flows, decode compressed action values, find record-trigger bindings.
- Investigations - pluggable analysis modules for stale automations, deprecated APIs, ACL conflicts, performance bottlenecks, and more.
- Service Catalog - browse catalogs, order items, manage carts.
Tool-package presets
Control the exposed surface with MCP_TOOL_PACKAGE. Four presets ship; custom comma-separated group lists are also accepted.
| Preset | Tools | Purpose |
|---|---|---|
full |
14 | Complete surface including writes and query builder |
readonly |
10 | Read operations, investigations, audit, and flow inspection |
core_readonly |
5 | Query, describe, and attachment only |
none |
1 | Only the list_tool_packages introspection tool |
Quickstart
# Install dependencies
uv sync
# Configure credentials
cp .env.example .env.local
# Edit .env.local with your instance URL, username, and password
# Run the server (stdio transport)
servicenow-platform-mcp
Three environment variables are required:
SERVICENOW_INSTANCE_URL- your instance (HTTPS enforced, e.g.https://dev12345.service-now.com)SERVICENOW_USERNAMESERVICENOW_PASSWORD
MCP client configuration
{
"mcpServers": {
"servicenow": {
"command": "servicenow-platform-mcp",
"transport": "stdio"
}
}
}
Safety posture
Eight sensitive tables (credential stores, OAuth entities, SSH keys) are unconditionally denied at the policy layer. Fields matching sensitive patterns - passwords, tokens, secrets, API keys, private keys, credentials - are masked in every response. Write operations use a preview-then-apply flow by default: the server returns a confirmation token that must be explicitly applied before any mutation reaches ServiceNow. In production environments (SERVICENOW_ENV=prod or production), all write operations are blocked at runtime regardless of which tool package is loaded.
Where to go next
- Getting Started - detailed setup and first-use walkthrough
- Configuration - full environment variable reference
- Tool Reference - per-tool parameter and response documentation
- Tool Packages - preset details and custom package composition
- Architecture - server bootstrap, decorator patterns, state management
- Safety and Policy - denied tables, masking, query safety, write gating
- Investigations - available analysis modules and their parameters
- Telemetry - opt-in Sentry integration
- Development - contributing, linting, testing
- INSTALL.md - alternative installation methods
License
MIT. See LICENSE.
Built by the ServiceNow MCP Contributors.
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。