Testing Bank MCP
An MCP router that connects to multiple PagerDuty sandbox environments for testing integrations like PD+Teams, PD+Jira, and more.
README
Testing Bank MCP
A Model Context Protocol (MCP) router that connects to multiple PagerDuty sandbox environments for testing integrations.
Architecture
┌─────────────────────────────────┐
│ Testing Bank Router MCP │
│ (Orchestrator/Gateway) │
└───────────────┬─────────────────┘
│
┌───────────┼───────────┐
│ │ │
▼ ▼ ▼
┌────────┐ ┌────────┐ ┌────────┐
│PD+Teams│ │PD+Jira │ │PD+Slack│ ... more environments
└────────┘ └────────┘ └────────┘
Available Environments
| Environment | Description |
|---|---|
pd-teams |
PagerDuty + Microsoft Teams |
pd-jira |
PagerDuty + Jira Cloud |
pd-salesforce |
PagerDuty + Salesforce |
pd-slack |
PagerDuty + Slack |
pd-aws |
PagerDuty + AWS CloudWatch/EventBridge |
Installation
cd testing-bank-mcp
npm install
Configuration
1. Set up environment files
Copy the example files and fill in your sandbox credentials:
# For PD + Teams
cp environments/pd-teams/.env.example environments/pd-teams/.env
# For PD + Jira
cp environments/pd-jira/.env.example environments/pd-jira/.env
2. Configure your MCP client
Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"testing-bank": {
"command": "node",
"args": ["router/index.js"],
"cwd": "/full/path/to/testing-bank-mcp"
}
}
}
Usage
Smart Routing (Recommended)
Just describe the customer issue and the router will detect the right environment:
"Customer reports PagerDuty incidents are not posting to their Teams channel"
The router will:
- Detect this is a PD + Teams issue
- Connect to the
pd-teamsenvironment - Give you access to all PD and Teams testing tools
Manual Environment Selection
1. List available environments: list_environments
2. Connect to specific environment: connect_environment("pd-jira")
3. Use environment tools: route_to_environment("pd-jira", "diagnose_integration")
Router Tools
| Tool | Description |
|---|---|
list_environments |
List all available testing environments |
detect_environment |
Auto-detect environment from issue description |
connect_environment |
Connect to a specific environment |
disconnect_environment |
Disconnect from an environment |
get_active_environments |
List currently connected environments |
route_to_environment |
Send commands to a connected environment |
smart_route |
Auto-detect and connect in one step |
Environment Tools (PD + Teams Example)
| Tool | Description |
|---|---|
pd_list_services |
List PagerDuty services |
pd_create_test_incident |
Create test incident |
pd_list_extensions |
List webhooks/integrations |
teams_send_test_message |
Send test message to Teams |
teams_verify_webhook |
Verify webhook configuration |
test_full_integration |
Run full end-to-end test |
diagnose_integration |
Diagnose common issues |
Example Workflow
User: "Customer says PD incidents aren't creating Jira tickets"
AI uses: smart_route("PD incidents aren't creating Jira tickets")
→ Detects: pd-jira environment
→ Connects to PD + Jira sandbox
AI uses: diagnose_integration()
→ Checks PD API connection
→ Checks Jira API connection
→ Lists Jira extensions in PD
→ Returns recommendations
AI uses: test_full_integration(service_id="PXXXXXX")
→ Creates test incident in PD
→ Waits for Jira sync
→ Searches for created ticket
→ Reports results
Adding New Environments
- Create directory:
environments/pd-newservice/ - Copy template from existing environment
- Modify API calls for the new service
- Add environment to
ENVIRONMENTSinrouter/index.js - Create
.env.examplewith required credentials
Security Notes
- Never commit
.envfiles - Use sandbox/test environments only
- Rotate API keys regularly
- Document which sandboxes are connected to production data (if any)
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。