Pipeline Assistant MCP
An MCP server that automates CI/CD pipeline creation, analysis, and security enforcement. It generates production-ready pipelines from templates and enforces DevSecOps best practices.
README
Pipeline Assistant MCP
AI-powered CI/CD pipeline automation using Model Context Protocol (MCP)
Transform pipeline creation from hours to seconds with guaranteed security compliance and DevSecOps best practices built-in.
What is Pipeline Assistant MCP?
Pipeline Assistant MCP is an intelligent system that automates the complete CI/CD pipeline lifecycle using AI. It leverages the Model Context Protocol (MCP) to provide context-aware pipeline generation, analysis, and improvement suggestions.
It's not just a validation tool - it's a complete DevSecOps assistant that:
- Generates production-ready pipelines from templates
- Enforces corporate security policies automatically
- Analyzes existing pipelines for vulnerabilities
- Provides actionable improvement suggestions
- Tracks compliance metrics across your organization
Why Pipeline Assistant MCP?
The Problem
Developer: "I need to create a pipeline for my .NET microservice"
2-4 hours later...
- Forgot security scanning stage
- Hardcoded database credentials
- Didn't configure dependency caching
- Tests don't generate coverage reports
- Deploys directly to production without approval
Result: Insecure, slow, non-compliant pipeline
The Solution
Developer: "Generate a .NET pipeline for production"
5 seconds later...
- Complete 6-stage pipeline generated
- All 10 security policies applied (SEC-001 to SEC-010)
- Optimized caching configured
- Tests with coverage reporting
- Production deployment with approval gates
- SBOM generation included
- Compliance Score: 98%
Result: Production-ready, secure, compliant pipeline
Business Value
| Metric | Before | After | Improvement |
|---|---|---|---|
| Pipeline creation time | 2-4 hours | 5 seconds | 99.9% faster |
| Security compliance | ~40% | 95%+ | +55% |
| Vulnerability detection | Manual review | Automatic | Real-time |
| Standards adoption | Inconsistent | Enforced | 100% coverage |
Architecture
System Overview
graph TB
subgraph "Developer Interfaces"
CLI[CLI Tools]
VSC[VS Code Extension]
CD[Claude Desktop]
GHA[GitHub Actions]
ADO[Azure DevOps]
end
subgraph "Core Services"
MCP[MCP Server]
PG[Pipeline Generator]
PA[Pipeline Analyzer]
PE[Policy Enforcer]
WM[Wiki Manager]
end
subgraph "Data Sources"
WIKI[Corporate Wiki v2.0]
POL[Security Policies]
TPL[Platform Templates]
MET[Adoption Metrics]
end
CLI --> MCP
VSC --> MCP
CD --> MCP
GHA --> MCP
ADO --> MCP
MCP --> PG
MCP --> PA
MCP --> PE
MCP --> WM
PG --> WIKI
PA --> POL
PE --> POL
WM --> MET
PG --> TPL
style MCP fill:#e1f5fe
style WIKI fill:#f3e5f5
style POL fill:#ffebee
Component Interaction
sequenceDiagram
participant D as Developer
participant M as MCP Server
participant G as Generator
participant E as Enforcer
participant W as Wiki
D->>M: Generate pipeline (dotnet, prod)
M->>W: Load standards v2.0
W-->>M: Stages, Policies, SLAs
M->>G: Create pipeline
G->>E: Apply security policies
E-->>G: SEC-001 to SEC-010
G-->>M: Complete pipeline
M-->>D: Pipeline + Compliance Score
Technology Stack
graph LR
subgraph "Runtime"
NODE[Node.js 20+]
TS[TypeScript 5.3]
end
subgraph "Protocol"
MCP[Model Context Protocol]
STDIO[STDIO Transport]
end
subgraph "Testing"
VIT[Vitest]
ZOD[Zod Validation]
end
subgraph "Integrations"
AZDO[Azure DevOps API]
GH[GitHub API]
VSCE[VS Code API]
end
NODE --> TS
TS --> MCP
MCP --> STDIO
TS --> VIT
TS --> ZOD
TS --> AZDO
TS --> GH
TS --> VSCE
Features
Core Capabilities
- Multi-Platform Support - Generate pipelines for Azure DevOps and GitHub Actions
- Pipeline Generation - Create complete pipelines from templates (.NET, Node.js, Python, Java, Go)
- Security Analysis - Detect hardcoded secrets, missing security stages, 15+ vulnerability types
- Policy Enforcement - Automatically apply SEC-001 to SEC-010 security policies
- Compliance Scoring - Calculate 0-100 scores with detailed breakdowns
- SBOM Generation - Software Bill of Materials for supply chain security
Integrations
- VS Code Extension - Real-time analysis, quick fixes, 35+ snippets
- Claude Desktop - Natural language pipeline generation via MCP
- GitHub Actions - Automatic PR analysis workflow
- Azure DevOps - PR Bot with webhook support
Security Features
- Webhook Signature Validation - HMAC-SHA256 with timing-safe comparison
- Secret Masking - Automatic redaction of tokens, passwords, API keys
- Rate Limiting - Sliding window algorithm to prevent abuse
- Input Validation - Zod schemas for all user inputs
Quick Start
Prerequisites
- Node.js 20+ and npm 9+
- Git
Installation
git clone https://github.com/soydachi/pipeline-assistant-mcp.git
cd pipeline-assistant-mcp
npm install
npm run build
npm test
Basic Usage
# Generate a pipeline for Azure DevOps
node dist/cli/pipeline-assistant.js generate \
--platform azure-devops \
--type dotnet \
--env production
# Generate a pipeline for GitHub Actions
node dist/cli/pipeline-assistant.js generate \
--platform github-actions \
--type node \
--env staging
# Analyze a pipeline
node dist/cli/pipeline-assistant.js analyze \
examples/pipelines/pipeline-con-problemas.yml
# List available platforms
node dist/cli/pipeline-assistant.js platforms
# List available templates
node dist/cli/pipeline-assistant.js templates --platform azure-devops
Project Structure
pipeline-assistant-mcp/
├── src/ # Core MCP server
│ ├── server.ts # MCP server entry point
│ ├── pipeline-generator.ts # Pipeline generation
│ ├── pipeline-analyzer.ts # Security analysis
│ ├── policy-enforcer.ts # Policy enforcement
│ ├── wiki-parser.ts # Standards parser
│ ├── wiki-manager.ts # Wiki management
│ ├── container.ts # Dependency injection
│ ├── platforms/ # Multi-platform support
│ │ ├── azure-devops.ts
│ │ └── github-actions.ts
│ ├── azure-devops/ # Azure DevOps integration
│ │ ├── client.ts
│ │ ├── pr-bot.ts
│ │ └── webhook-handler.ts
│ └── utils/ # Shared utilities
│ ├── logger.ts
│ ├── validation.ts
│ └── rate-limiter.ts
├── cli/ # Command-line tools
│ ├── pipeline-assistant.ts
│ ├── wiki-cli.ts
│ └── pr-bot-cli.ts
├── vscode-extension/ # VS Code extension
├── wiki/standards/ # Corporate standards v2.0
│ ├── core/ # Stage definitions
│ ├── security/ # Security policies
│ ├── quality/ # Quality gates
│ ├── platforms/ # Platform templates
│ │ ├── azure/templates/
│ │ └── github/templates/
│ ├── migration/ # Migration guides
│ └── governance/ # Governance docs
├── tests/ # Test suite (341+ tests)
└── examples/ # Example pipelines
Documentation
| Document | Description |
|---|---|
| Workshop Guide | Complete tutorial with architecture deep-dive |
| Usage Guide | Reference for all platforms and configurations |
| Contributing | How to contribute to the project |
| Changelog | Version history and release notes |
Integrations
MCP Server (Claude Desktop)
{
"mcpServers": {
"pipeline-assistant": {
"command": "node",
"args": ["dist/src/server.js"],
"cwd": "/path/to/pipeline-assistant-mcp"
}
}
}
VS Code Extension
cd vscode-extension
npm install && npm run compile
# Press F5 to launch in development mode
Azure DevOps
export AZDO_ORG_URL="https://dev.azure.com/your-org"
export AZDO_PAT="your-personal-access-token"
export AZDO_PROJECT="your-project"
GitHub Actions
Add .github/workflows/pipeline-review.yml to automatically analyze PRs.
See Usage Guide for detailed configuration.
Standards v2.0
Pipeline Assistant uses a structured standards system:
Security Policies (SEC-001 to SEC-010)
| Policy | Name | Level |
|---|---|---|
| SEC-001 | Secret Scanning | Mandatory |
| SEC-002 | SAST Analysis | Mandatory |
| SEC-003 | Dependency Scanning | Mandatory |
| SEC-004 | Container Scanning | Conditional |
| SEC-007 | DAST | Conditional |
| SEC-008 | License Compliance | Mandatory |
| SEC-010 | SBOM Generation | Mandatory |
Mandatory Pipeline Stages
- Validate - Linting, formatting, type checking
- Security - All security scans (parallel)
- Build - Application build + SBOM
- Test - Unit + Integration tests
- Scan - Container security
- Deploy - Environment deployments
Development
npm run dev # Watch mode
npm test # Run tests (341+ tests)
npm run lint # Check code style
npm run build # Build project
Testing
# Run all tests
npm test
# Run specific test
npx vitest run tests/policy-enforcer.test.ts
# Run with coverage
npx vitest run --coverage
Contributing
We welcome contributions! Please see CONTRIBUTING.md for guidelines.
License
Author
Dachi Gogotchuri (@soydachi)
- Website: soydachi.com
- LinkedIn: Dachi Gogotchuri
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。