SynapseQL

The Ultimate AI-Powered Federated Query Builder

Overview

This server provides a flexible API for querying data across multiple PostgreSQL databases.. It supports:

• Standard queries with filtering, sorting and pagination
• Relation-based queries with automatic joins
• Cross-database joins
• Health monitoring
• Raw SQL execution (for admin use)
• Fallback in case on instance goes down

The server uses a schema definition file (schema.json) to validate all queries against the defined database structure.

Connection Management

The server automatically:

• Creates connection pools for all databases defined in schema.json.
• Tests connections at startup
• Maintains connection pooling settings (timeout, max connections)
• Handles graceful shutdown of database connections

API Endpoints

1. Standard Query: /query (POST)

The main endpoint for querying data from tables with support for filtering, relations, and pagination.

Request Format

{
  "tableName": {
    "select": ["field1", "field2"],
    "field1": "value1",
    "field2": { "gt": 100 },
    "orderBy": "field1",
    "limit": 10,
    "offset": 0,
    "groupBy": ["field1"],
    "relations": {
      "relatedTable": {
        "select": ["rel_field1", "rel_field2"],
        "orderBy": { "rel_field1": "DESC" },
        "limit": 5,
        "relations": {
          "nestedRelation": {
            "select": ["nested_field1"]
          }
        }
      }
    }
  }
}

Query Options

Option	Type	Description
select	string[]	Columns to select (defaults to all columns if omitted)
orderBy	string | string[] | object	Sorting options (can be a single field, array of fields, or object with field/direction)
limit	number	Maximum number of records to return
offset	number	Number of records to skip
groupBy	string[]	Fields to group results by
relations	object	Define related data to load with the query

Filter Operators

Operator	Example	Description
= (default)	"field": "value"	Exact match
gt	"field": { "gt": 100 }	Greater than
lt	"field": { "lt": 100 }	Less than
gte	"field": { "gte": 100 }	Greater than or equal
lte	"field": { "lte": 100 }	Less than or equal
like	"field": { "like": "%text%" }	LIKE pattern match (case-sensitive)
ilike	"field": { "ilike": "%text%" }	ILIKE pattern match (case-insensitive)
neq	"field": { "neq": "value" }	Not equal to
in	"field": { "in": [1, 2, 3] }	Value is in array
notIn	"field": { "notIn": [1, 2, 3] }	Value is not in array
between	"field": { "between": [1, 10] }	Value is between two values (inclusive)
isNull	"field": { "isNull": true }	Field is NULL
isNotNull	"field": { "isNotNull": true }	Field is NOT NULL

Response Format

{
  "tableName": [
    {
      "field1": "value1",
      "field2": "value2",
      "relatedTable": [
        {
          "rel_field1": "value",
          "rel_field2": "value",
          "nestedRelation": [
            {
              "nested_field1": "value"
            }
          ]
        }
      ]
    }
  ]
}

2. Cross-Database Join: /join (POST)

Performs joins across tables that may exist in different databases.

Request Format

{
  "mainTable": "users",
  "mainSelect": ["id", "name", "email"],
  "mainFilters": {
    "status": "active"
  },
  "joins": [
    {
      "table": "orders",
      "select": ["id", "order_date", "total"],
      "localKey": "id",
      "foreignKey": "user_id",
      "filters": {
        "status": "completed"
      }
    },
    {
      "table": "payments",
      "select": ["id", "amount", "date"],
      "localKey": "id",
      "foreignKey": "user_id",
      "filters": {
        "status": "processed"
      }
    }
  ]
}

Join Options

Option	Type	Description
mainTable	string	Primary table to query
mainSelect	string[]	Fields to select from main table
mainFilters	object	Filters to apply to main table
joins	array	Array of tables to join
joins[].table	string	Related table name
joins[].select	string[]	Fields to select from related table
joins[].localKey	string	Field in main table for join
joins[].foreignKey	string	Field in related table for join
joins[].filters	object	Additional filters for related table

Response Format

[
  {
    "id": 1,
    "name": "User Name",
    "email": "[email protected]",
    "orders": [
      {
        "id": 101,
        "order_date": "2023-01-15T00:00:00.000Z",
        "total": 99.99
      }
    ],
    "payments": [
      {
        "id": 201,
        "amount": 99.99,
        "date": "2023-01-15T00:00:00.000Z"
      }
    ]
  }
]

3. Raw SQL: /raw-sql (POST)

Execute raw SQL queries directly against a database (admin use only).

Request Format

{
  "db": "database_name",
  "sql": "SELECT * FROM users WHERE id = $1",
  "params": [123]
}

Options

Option	Type	Description
db	string	Database name to execute against (must match schema.json)
sql	string	Raw SQL query to execute
params	array	Optional parameters for prepared statement

Response Format

{
  "rowCount": 1,
  "rows": [
    {
      "id": 123,
      "name": "Example User",
      "email": "[email protected]"
    }
  ]
}

4. Health Check: /health (GET)

Check the health and connection status of all configured databases.

Response Format

{
  "status": {
    "database1": "connected",
    "database2": "connected"
  },
  "message": "All systems operational"
}

Schema Definition

The server validates all queries against a schema defined in schema.json. This file must define:

1. Database connections
2. Table definitions
3. Column definitions
4. Relationship definitions

Example schema structure:

{
  "databases": {
    "main_db": {
      "host": "localhost",
      "port": 5432,
      "user": "postgres",
      "password": "password",
      "database": "main_database"
    },
    "analytics_db": {
      "host": "localhost",
      "port": 5432,
      "user": "postgres",
      "password": "password",
      "database": "analytics_database"
    }
  },
  "tables": {
    "users": {
      "db": "main_db",
      "columns": ["id", "name", "email", "created_at"],
      "relations": {
        "orders": {
          "table": "orders",
          "foreign_key": "id",
          "reference": "user_id"
        }
      }
    },
    "orders": {
      "db": "main_db",
      "columns": ["id", "user_id", "total", "status", "created_at"],
      "relations": {
        "items": {
          "table": "order_items",
          "foreign_key": "id",
          "reference": "order_id"
        }
      }
    },
    "order_items": {
      "db": "main_db",
      "columns": ["id", "order_id", "product_id", "quantity", "price"]
    },
    "user_metrics": {
      "db": "analytics_db",
      "columns": ["user_id", "logins_count", "last_active", "engagement_score"]
    }
  }
}

Error Handling

The server returns appropriate HTTP status codes and detailed error messages:

• 400 - Bad Request: Invalid query structure or parameters
• 404 - Not Found: Requested resource not found
• 500 - Internal Server Error: Unexpected server error
• 503 - Service Unavailable: Database connection issues

Error responses include:

• Error message
• Stack trace (in development mode only)

Query Examples

Basic Query with Filtering

{
  "users": {
    "select": ["id", "name", "email"],
    "status": "active",
    "created_at": { "gte": "2023-01-01" },
    "limit": 10
  }
}

Query with Relations

{
  "users": {
    "id": 123,
    "relations": {
      "orders": {
        "select": ["id", "total", "created_at"],
        "status": "completed",
        "orderBy": { "created_at": "DESC" },
        "limit": 5
      }
    }
  }
}

Complex Query with Nested Relations

{
  "users": {
    "role": "customer",
    "relations": {
      "orders": {
        "status": "completed",
        "created_at": { "gte": "2023-01-01" },
        "relations": {
          "items": {
            "select": ["product_id", "quantity", "price"],
            "quantity": { "gt": 1 }
          }
        }
      }
    }
  }
}

Cross-Database Join

{
  "mainTable": "users",
  "mainSelect": ["id", "name", "email"],
  "mainFilters": {
    "status": "active"
  },
  "joins": [
    {
      "table": "user_metrics",
      "select": ["logins_count", "engagement_score"],
      "localKey": "id",
      "foreignKey": "user_id",
      "filters": {
        "engagement_score": { "gt": 50 }
      }
    }
  ]
}