Harper MCP Server
Exposes HarperDB tables and custom resources as structured MCP resources, enabling read-only access and filtering through JSON-RPC calls.
README
Harper MCP Server
A server implementation of the Model Context Protocol (MCP), designed to expose data in HarperDB as structured "Resources" accessible via standardized JSON-RPC calls.
Note: Requires HarperDB version 4.5.10 or later.
What is Harper
Harper is a Composable Application Platform that merges database, cache, app logic, and messaging into a single runtime. Components like this plug directly into Harper, letting you build and scale distributed services fast, without managing separate systems. Built for geo-distributed apps with low latency and high uptime by default.
Features
- MCP-compatible API server for Harper
- Predefined static capabilities endpoint (
/capabilities.json) - Provides read-only access to data from Harper tables and custom resources
- Supports filtering data in Harper tables using query parameters
- Handles pagination (limit, start) for table data retrieval
- Provides standardized error responses
Getting Started
Prerequisites
- Harper stack installed globally.
- Ensure HarperDB v4.5.10 or later is configured and running with necessary databases and schemas.
- Environment variable
HOSTshould be set to the base URL of your server. This is used to construct resource URIs.
Deploying to Harper
The Harper mcp-server is published to NPM and can be installed using Harper's Operation API.
i.e.
POST https://harper-server.com:9925
{
"operation": "deploy_component",
"package": "@harperdb/mcp-server@1.0.0"
}
Security & Authentication
Harper employs role-based, attribute-level security to ensure users access only authorized data. Requests to the server are authenticated using HarperDB's built-in authentication mechanisms, which include Basic Auth, JWT, and mTLS. See Harper Security Docs for more details.
API
MCP Methods
The server implements the following MCP methods:
resources/list: Lists all available resources (HarperDB tables and custom resources).resources/read: Retrieves data for a specific resource based on its URI.
A single endpoint, /mcp handles all requests. The server uses JSON-RPC 2.0 for communication.
- Request Format: All requests are sent as JSON-RPC 2.0 formatted JSON objects.
- Response Format: The server responds with JSON-RPC 2.0 formatted JSON objects.
- Error Handling: The server returns standardized error responses.
Resource URIs
-
Tables: Resources representing HarperDB tables are accessed via URIs like:
{HOST}/{table_name}- Example:
http://localhost:9925/my_table
- Example:
-
Table Rows: Individual rows within a table can be accessed using the primary key:
{HOST}/{table_name}/{primary_key_value}- Example:
http://localhost:9925/my_table/123(where 123 is the primary key value)
- Example:
-
Custom Resources: Custom resources are accessed via URIs defined by their registered path:
{HOST}/{path}/{resource_name}- Example:
http://localhost:9925/custom/my_resource
- Example:
Usage
1. Listing Resources
POST /mcp
Sample Request:
{
"jsonrpc": "2.0",
"id": 1,
"method": "resources/list"
}
Sample Response:
{
"jsonrpc": "2.0",
"id": 1,
"result": {
"resources": [
{
"uri": "http://localhost:9926/CustomerOrders",
"name": "CustomerOrders",
"description": "CustomerOrders table with attributes: id (PK - Int), customerId (string), customer (Relationship from customerId - Customers), itemSku (String), item (Relationship from itemSku - Items), subTotal (Float), orderTotal (Float), date (DateTime). Results can be filtered with optional query parameters.",
"mimeType": "application/json"
},
{
"uri": "http://localhost:9926/Customers",
"name": "Customers",
"description": "Customers table with attributes: id (PK - Int), email (String), phoneNumber (String), customerName (String), country (String), orders (Relationship to customerId - array). Results can be filtered with optional query parameters.",
"mimeType": "application/json"
},
{
"uri": "http://localhost:9926/Items",
"name": "Items",
"description": "Items table with attributes: sku (PK - String), itemName (String), itemDescription (String), unitPrice (Int). Results can be filtered with optional query parameters.",
"mimeType": "application/json"
},
{
"uri": "http://localhost:9926/TestCustomer",
"name": "TestCustomer",
"description": "Customers table with attributes: id (PK - Int), email (String), phoneNumber (String), customerName (String), country (String), orders (Relationship to customerId - array). Results can be filtered with optional query parameters.",
"mimeType": "application/json"
},
{
"uri": "http://localhost:9926/api/test",
"name": "test",
"description": "REST Resource.",
"mimeType": "application/json"
}
]
}
}
2. Get resources data
POST /mcp
Sample Request:
{
"jsonrpc": "2.0",
"id": 2,
"method": "resources/read",
"params": {
"uri": "http://localhost:9926/Customers"
}
}
Sample Response:
{
"jsonrpc": "2.0",
"id": 2,
"result": {
"contents": [
{
"uri": "http://localhost:9926/Customers/11",
"mimeType": "application/json",
"text": "{\"id\":11,\"email\":\"kelly.williams@example.com\",\"phoneNumber\":\"214-555-1234\",\"customerName\":\"Kelly Williams\",\"country\":\"USA\"}"
},
{
"uri": "http://localhost:9926/Customers/12",
"mimeType": "application/json",
"text": "{\"id\":12,\"email\":\"liam.martinez@example.com\",\"phoneNumber\":\"972-555-5678\",\"customerName\":\"Liam Martinez\",\"country\":\"Canada\"}"
},
{
"uri": "http://localhost:9926/Customers/13",
"mimeType": "application/json",
"text": "{\"id\":13,\"email\":\"mia.anderson@example.com\",\"phoneNumber\":\"469-555-9012\",\"customerName\":\"Mia Anderson\",\"country\":\"UK\"}"
},
{
"uri": "http://localhost:9926/Customers/14",
"mimeType": "application/json",
"text": "{\"id\":14,\"email\":\"noah.thomas@example.com\",\"phoneNumber\":\"817-555-3456\",\"customerName\":\"Noah Thomas\",\"country\":\"Australia\"}"
},
{
"uri": "http://localhost:9926/Customers/15",
"mimeType": "application/json",
"text": "{\"id\":15,\"email\":\"olivia.jackson@example.com\",\"phoneNumber\":\"682-555-7890\",\"customerName\":\"Olivia Jackson\",\"country\":\"Germany\"}"
},
{
"uri": "http://localhost:9926/Customers/16",
"mimeType": "application/json",
"text": "{\"id\":16,\"email\":\"owen.white@example.com\",\"phoneNumber\":\"214-555-2345\",\"customerName\":\"Owen White\",\"country\":\"France\"}"
},
{
"uri": "http://localhost:9926/Customers/17",
"mimeType": "application/json",
"text": "{\"id\":17,\"email\":\"sophia.harris@example.com\",\"phoneNumber\":\"972-555-6789\",\"customerName\":\"Sophia Harris\",\"country\":\"Japan\"}"
}
]
}
}
Querying Tables
When retrieving data from tables using resources/read, you can use optional query parameters in the URI to filter data.
- Filtering: Use
attribute=valuepairs to filter based on column values. The comparator is always "equals".- Example:
http://localhost:9925/my_table?name=John&city=NewYork
- Example:
- Pagination: Use
limitandstartparameters for pagination.limit: Maximum number of results to return.start: Offset to start returning results from.- Example:
http://localhost:9925/my_table?limit=10&start=20
Error Responses
The server returns standardized JSON-RPC error responses:
{
"jsonrpc": "2.0",
"id": 2,
"error": {
"code": -32602,
"message": "Invalid params."
}
}
Error Codes:
-32601: Method not found.-32602: Invalid params.-32603: Internal server error.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.