sentinel-mcp
Enables local AI-powered API testing using Ollama as the underlying AI provider, allowing users to run tests with local models like llama3 for privacy and offline testing.
Provides AI-powered API testing capabilities using OpenAI as the underlying AI provider, enabling comprehensive API testing with OpenAI's language models.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@sentinel-mcptest my login endpoint with security checks using openai"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
sentinel-mcp
MCP Server untuk AI-powered API testing — positive, negative, dan security testing langsung dari chat AI.
Fitur
3 jenis testing: positive (happy path), negative (edge case), security (SQLi, XSS, auth, dll)
Multi AI provider: Ollama (local), Claude, OpenAI, Gemini — bisa switch per perintah
Auto report: hasil analisis AI langsung disimpan sebagai Markdown
MCP native: langsung dipakai dari Claude Desktop, Cursor, Windsurf, Cline
Struktur Project
sentinel-mcp/
├── src/
│ ├── index.ts # MCP server entry point
│ ├── types.ts # Type definitions
│ ├── tools/
│ │ └── apiTest.ts # MCP tool handlers
│ ├── runner/
│ │ ├── runner.ts # HTTP test engine
│ │ └── loader.ts # YAML config loader
│ ├── providers/
│ │ ├── factory.ts # Provider resolver
│ │ ├── prompt.ts # Shared prompt builder
│ │ ├── ollama.ts # Ollama adapter
│ │ ├── claude.ts # Claude adapter
│ │ ├── openai.ts # OpenAI adapter
│ │ └── gemini.ts # Gemini adapter
│ └── reports/
│ └── generator.ts # Markdown report generator
├── tests/endpoints/
│ └── suite.yaml # Definisi endpoint yang ditest
├── reports/ # Output report (auto-created)
├── .env.example
└── claude_desktop_config.example.jsonQuick Start
1. Install dependencies
cd sentinel-mcp
npm install2. Setup environment
cp .env.example .env
# Edit .env sesuai kebutuhanmu3. Install Ollama (untuk local AI)
# macOS / Linux
curl -fsSL https://ollama.ai/install.sh | sh
ollama pull llama34. Definisikan endpoint di suite.yaml
Secara default, test config dimuat dari suites/, tests/endpoints/suite.yaml, atau lokasi default lainnya.
Edit file tersebut sesuai API kamu. Jika kamu memiliki test suite di luar project, kamu bisa memberikan path langsung ke MCP via chat AI (mendukung single file config maupun satu directory khusus berisikan banyak konfigurasi).
5. Daftarkan ke Claude Desktop
Salin isi claude_desktop_config.example.json ke:
macOS:
~/Library/Application Support/Claude/claude_desktop_config.jsonWindows:
%APPDATA%\Claude\claude_desktop_config.json
Ganti path absolut sesuai lokasi project kamu, lalu restart Claude Desktop.
Cara Pakai di Chat (Contoh Prompt AI)
Kamu dapat menggunakan bahasa manusia secara natural di AI client yang mendukung MCP (seperti Claude Desktop, Cline, Windsurf). Di balik layar, AI akan otomatis memilih tool yang relevan.
1. Basic Testing (Menggunakan file config default):
"Tolong jalankan API test untuk semua endpoint saya" "Test endpoint login saja pakai gemini ya" "Coba run security test untuk sistem saya"
2. Menggunakan File / Direktori Eksternal (suite_dir / suite_file):
Jika konfigurasi test API kamu ada di folder project lain (di luar project Sentinel MCP), kamu tinggal beri tahu path-nya ke AI:
"Test API saya menggunakan config di folder
/Users/nama/ProjectLain/api-suites/" "Tolong jalankan security test berdasarkan file/path/to/custom-project/suite.yaml"
3. Kombinasi:
"Bantu cek security untuk API payment saya yang ada di
/Users/dev/ecommerce/suites/, gunakan provider klaude karena saya butuh analisa security tingkat tinggi."
4. Mengatur Provider AI:
"Provider apa saja yang saat ini tersedia?" "Bisa switch ke openai untuk test endpoint register?"
Security Testing yang Dilakukan
Test | Deskripsi |
No Auth | Akses endpoint tanpa Authorization header |
SQL Injection | Payload |
XSS | Payload |
Info Leak | Cek apakah error response bocorkan stack trace / SQL |
Oversized Payload | Kirim 100KB payload untuk cek DoS resistance |
Format Report
Report disimpan di folder reports/ dengan nama YYYY-MM-DD-[provider].md.
Isi report:
Skor keseluruhan (0-100)
Ringkasan dari AI
Daftar bug dengan severity dan rekomendasi fix
Temuan keamanan dengan cara mitigasi
Detail per endpoint (semua test case)
Menambah Endpoint Baru
Edit tests/endpoints/suite.yaml:
- name: nama-endpoint
method: POST # GET | POST | PUT | PATCH | DELETE
path: /api/resource
auth: true # sertakan Authorization header?
expectedStatus: 201
body:
field: value
expectedFields:
- id
- nameThis server cannot be installed
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/kal72/sentinel-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server