Skip to content

operationDetails_doc.md

Latest commit

 

History

History
151 lines (127 loc) · 3.3 KB

operationDetails_doc.md

File metadata and controls

151 lines (127 loc) · 3.3 KB

operationDetails Parameter

The operationDetails parameter allows you to fine-tune query operations with operational directives similar to GraphQL capabilities. It accepts a JSON array containing one or more operation directives that refine the shape and scope of returned objects.

Parameter Format

[
  {
    "opType": "<operation_type>",
    "<operation_specific_details>": [...]
  }
]

Supported Operation Types

1. projections - Select Specific Attributes

Specify which attributes to include for particular object types (similar to GraphQL field selection).

Format:

{
  "opType": "projections",
  "projectionsDetails": [
    {
      "type": "<ClassName>",
      "attribs": ["attribute1", "attribute2", "..."]
    }
  ]
}

Example: Only retrieve name, id, and email from Employee objects:

[{
  "opType": "projections", 
  "projectionsDetails": [{
    "type": "Employee", 
    "attribs": ["name", "id", "email"]
  }]
}]

2. ignore - Skip Object References

Exclude specific referenced objects from deep queries (useful for avoiding circular references or unnecessary data).

Format:

{
  "opType": "ignore",
  "references": ["<ClassName>", "<attributeName>"]
}

Example: Ignore the departments reference in Employee objects:

[{
  "opType": "ignore",
  "references": ["Employee", "departments"]
}]

3. follow - Include Object References

Force inclusion of specific referenced objects in shallow queries.

Format:

{
  "opType": "follow", 
  "references": ["<ClassName>", "<attributeName>"]
}

Example: Include the manager reference in Employee objects:

[{
  "opType": "follow",
  "references": ["Employee", "manager"] 
}]

4. filters - Apply Predicates

Filter objects of specific types using predicate expressions.

Format:

{
  "opType": "filters",
  "predicates": [
    {
      "type": "<ClassName>",
      "predicate": "<filter_expression>"
    }
  ]
}

Example: Only include active employees with salary > 50000:

[{
  "opType": "filters",
  "predicates": [{
    "type": "Employee",
    "predicate": "active = true AND salary > 50000"
  }]
}]

Combining Multiple Operations

You can combine multiple operation types in a single operationDetails array:

[
  {
    "opType": "projections",
    "projectionsDetails": [{
      "type": "Employee", 
      "attribs": ["name", "id", "salary"]
    }]
  },
  {
    "opType": "filters", 
    "predicates": [{
      "type": "Employee",
      "predicate": "department = 'Engineering'"
    }]
  },
  {
    "opType": "ignore",
    "references": ["Employee", "personalDetails"]
  }
]

Usage Tips

  • Empty string or null: No operational directives applied
  • projections: Most useful for reducing payload size and improving performance
  • ignore/follow: Control object graph traversal depth and scope
  • filters: Apply server-side filtering before data transmission
  • Combine operations: Mix and match operations for precise data control

MCP Client Implementation

When calling the query tool, pass the operationDetails as either:

  • JSON object/array (if your MCP client auto-parses JSON)
  • JSON string (if your MCP client expects string parameters)

The MCP server will automatically handle both formats.