Provides tools to interact with Stripe API, enabling AI agents to perform operations like listing charges securely with credential injection.
Click on "Deploy 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., "@Agent Data Gateway Get recent charges from Stripe"
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.
Open-core middleware that gives AI agents secure, auditable access to production data. Converts existing APIs/DBs into MCP tools with credential hiding, data scoping, and immutable audit logs.
┌─────────────┐ ┌──────────────────────────────────┐ ┌──────────────┐
│ Agent │────▶│ ADG Proxy (Fastify) │────▶│ Backend │
│ (no creds) │ │ ┌────────────┐ ┌─────────────┐ │ │ API / DB │
│ │ │ │ Interceptor │──│ Credential │ │ │ │
└─────────────┘ │ │ inject creds│ │ Store (mem) │ │ └──────────────┘
│ └────────────┘ └─────────────┘ │
│ ┌────────────┐ │
│ │ Audit Log │ │
│ └────────────┘ │
└──────────────────────────────────┘Related MCP server: heddle
# Install
npm install
# Build
npm run build
# Configure
cp examples/config.yaml config.yaml
# Edit config.yaml with your sources
# Run
npm start # production
npm run dev # development with hot reloadPOST /proxyAgent sends a request referencing a source by name. Proxy injects credentials server-side.
{
"source": "stripe-api",
"url": "https://api.stripe.com/v1/charges",
"method": "GET"
}GET /healthHealth check — returns status and source count.
GET /sourcesLists known source names (no credentials).
Credentials stored in-memory only, never persisted to disk after boot
All credential fields redacted from logs, error messages, and responses
CredentialNotFoundError leaks only the source name, never credential values
Credential store is immutable after initialization
├── src/
│ ├── index.ts # Entry point & Fastify server
│ ├── config/
│ │ ├── loader.ts # YAML config loader
│ │ └── schema.ts # Zod schema & types
│ ├── credentials/
│ │ ├── store.ts # In-memory credential store
│ │ ├── types.ts # Credential types
│ │ └── redact.ts # Credential redaction utilities
│ ├── proxy/
│ │ └── interceptor.ts # Credential injection middleware
│ └── audit/
│ └── logger.ts # Immutable audit logger
├── tests/
│ ├── credentials/
│ │ ├── redact.test.ts
│ │ └── store.test.ts
│ └── proxy/
│ └── interceptor.test.ts
├── examples/
│ └── config.yaml
├── docker-compose.yml
└── package.jsonThis server cannot be deployed
Governed MCP gateway: one endpoint for your tools, with credential custody and audit log.
One MCP URL for all your connectors — scoped writes, enforced constraints, and a full audit trail.
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
Zero-secret MCP gateway for AI agents: risk-scored, audited calls with human-in-the-loop approval.
