jenkins-slack-mcp
Enables triggering Jenkins builds from IDEs with Slack notifications, requiring only a one-time login.
README
jenkins-slack-mcp
MCP server to trigger Jenkins builds from any IDE with Slack DM notifications. Auto-discovers all jobs on login — no manual configuration.
Install
npm install -g jenkins-slack-mcp
Register in your IDE
Amazon Q
~/.aws/amazonq/mcp.json:
{
"mcpServers": {
"jenkins-slack": {
"command": "jenkins-slack-mcp",
"args": [],
"disabled": false
}
}
}
VS Code
.vscode/mcp.json:
{
"mcpServers": {
"jenkins-slack": { "command": "jenkins-slack-mcp" }
}
}
Claude Desktop
~/Library/Application Support/Claude/claude_desktop_config.json (macOS):
{
"mcpServers": {
"jenkins-slack": { "command": "jenkins-slack-mcp" }
}
}
Cursor
.cursor/mcp.json:
{
"mcpServers": {
"jenkins-slack": { "command": "jenkins-slack-mcp" }
}
}
Without global install
{
"mcpServers": {
"jenkins-slack": {
"command": "npx",
"args": ["-y", "jenkins-slack-mcp"]
}
}
}
Flow of Work
┌─────────────────────────────────────────────────────────┐
│ 1. login_jenkins (baseUrl only) │
│ → Opens browser to Jenkins token page │
│ → User copies API token + build trigger token │
│ │
│ 2. login_jenkins (all params) │
│ → Validates credentials │
│ → Auto-discovers ALL jobs from Jenkins │
│ → Stores encrypted at ~/.jenkins-slack-mcp/ │
│ │
│ 3. list_jobs / list_jobs filter="tms" │
│ → Shows jobs in table with status (✅ ❌ ⏸️) │
│ │
│ 4. job_details jobName="my-job" │
│ → Shows all parameters (name, type, default, │
│ choices) in table format │
│ │
│ 5. trigger_build jobName="my-job" params={...} │
│ → Triggers build with params │
│ → Sends Slack DM to you (if configured) │
│ → Posts to channel (if notifyChannel provided) │
└─────────────────────────────────────────────────────────┘
Usage
Step 1: Login to Jenkins
First time — just provide the URL, browser opens automatically:
login_jenkins baseUrl="https://jenkins.example.com"
→ Browser opens Jenkins → Copy your API token and build trigger token.
Then login with full credentials:
login_jenkins:
baseUrl: https://jenkins.example.com
user: your_username
apiToken: your_api_token
buildToken: your_build_trigger_token
→ All jobs auto-discovered!
Step 2: Browse Jobs
list_jobs # all jobs
list_jobs filter="tms" # filter by name
Output:
| # | Job Name | Status |
|---|---|---|
| 1 | build-main | ✅ Success |
| 2 | lib-packages | ✅ Success |
| 3 | job-constant | ✅ Success |
Step 3: Check Parameters
job_details jobName="tms-docker-build-new"
Output:
| Parameter | Type | Default | Choices | Description |
|---|---|---|---|---|
| BRANCH | String | main | - | Branch to build |
Step 4: Trigger Build
trigger_build jobName="tms-docker-build-new" params={"BRANCH": "feature/xyz"}
Step 5 (Optional): Setup Slack DMs
setup_slack:
botToken: xoxb-your-bot-token
userId: U0123456789
How to get your Slack User ID:
- Open Slack → Click your profile picture
- Click "Profile" → Click "..." (more)
- Click "Copy Member ID"
How to get bot token:
- Go to https://api.slack.com/apps → Your App
- OAuth & Permissions → Bot User OAuth Token (starts with
xoxb-)
After setup, every trigger_build sends you a DM automatically.
Available Tools
| Tool | Description |
|---|---|
login_jenkins |
Login + auto-discover jobs (opens browser for new users) |
setup_slack |
Configure Slack bot token + User ID for DM notifications |
status |
Check login status + job count |
list_jobs |
All jobs in table format (filterable) |
job_details |
Show parameters for a job |
trigger_build |
Trigger build + Slack DM + channel notify |
refresh_jobs |
Re-fetch jobs from Jenkins |
whoami |
Current user details |
logout |
Clear all credentials |
Security
BuildPilot follows OWASP security best practices:
Credential Storage
- All credentials (API tokens, Slack tokens) are encrypted at rest using AES-256-GCM
- Encryption key is machine-bound (derived from hostname + user identity)
- Config file stored with
600permissions (owner-read/write only) - Atomic file writes prevent data corruption and race conditions
Network Security
- SSRF Protection — Private IPs, loopback, link-local, and cloud metadata endpoints (169.254.169.254) are blocked
- TLS enforced — Only HTTP/HTTPS protocols allowed
- Request timeouts — All HTTP calls have strict timeouts (10-30s)
- Redirect protection — Auth headers stripped on redirects
Input Validation
- Build parameter names validated against allowlist (
[a-zA-Z0-9_\-.]) - Reserved parameter keys (
token,cause,json,submit) are blocked to prevent injection - Parameter values capped at 1000 characters
- Jenkins URL validated and sanitized before use
OAuth Security
- OAuth callback server binds to
127.0.0.1only (not0.0.0.0) - CSRF state parameter uses
crypto.randomBytes(32) - Timing-safe comparison for state validation
- Rate limiting on callback endpoint (max 5 attempts)
- Auto-timeout after 2 minutes
- All non-callback routes rejected with 404
Error Handling
- Error messages sanitized to prevent credential leakage
- URLs, auth headers, and tokens are redacted from error output
- Error messages truncated to 200 characters
VS Code Extension
See vscode-extension/README.md for the full-featured VS Code extension with:
- Inline search, grouped jobs, pin/unpin
- Live build summary webview
- Stop running builds from history
- Slack notifications
Credentials
Stored encrypted at ~/.jenkins-slack-mcp/config.enc (AES-256-GCM, 600 permissions, owner-only).
Author
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。