v0.1.4 is here!

[CheMiguel23]

Key Improvements:

• Consistent Error Handling: Implemented consistent error handling across all tool handlers, adhering to MCP guidelines for tool errors (using isError: true and the content array).
• AI-Friendly Responses: Formatted error responses to be more easily understood by AI models, providing context, suggestions, and potential recovery steps.
• Centralized Configuration: Moved configuration variables to config/config.js for better organization and maintainability.
• Improved Readability and Maintainability: Used more descriptive variable names and added comments to improve code clarity.

[CheMiguel23]

Example: update_spaceship Error

When an error occurs during a tool call, MemoryMesh returns a structured error response that includes details about the error, the operation that was attempted, the context of the error, and suggestions for resolution.

Scenario: The AI attempts to update a spaceship using the update_spaceship tool, but the specified spaceship node does not exist in the knowledge graph.

Input:

{
  "update_spaceship": {
    "name": "Nebula Voyager-2",
    "Hull Integrity": "100%"
  }
}

Error Response:

{
  "isError": true,
  "content": [
    {
      "type": "text",
      "text": "Error during update_spaceship: Error: update_spaceship \"Nebula Voyager-2\" not found"
    },
    {
      "type": "text",
      "text": "Context: {\"input\":{\"update_spaceship\":{\"name\":\"Nebula Voyager-2\",\"Hull Integrity\":\"100%\"}}}"
    },
    {
      "type": "text",
      "text": "Suggestions: [\"Ensure a add_spaceship with the name \\\"Nebula Voyager-2\\\" exists\", \"Review the error message and the provided arguments\", \"Check the schema definition for required fields and types\"]"
    },
    {
      "type": "text",
      "text": "Recovery Steps: [\"Modify the input to match the schema requirements\", \"If the error is related to data, consider correcting the data and retrying\"]"
    }
  ],
  "timestamp": "2024-12-12T05:30:00.000Z"
}

How this helps the AI:

1. Clear Error Indication: The isError: true flag immediately signals to the AI that the operation failed.
2. Specific Error Message: The error message pinpoints the problem (node not found).
3. Contextual Information: The Context field shows the exact input that caused the error, allowing the AI to cross-reference with its understanding of the knowledge graph.
4. Actionable Suggestions: The Suggestions provide guidance on how to potentially resolve the issue (e.g., checking if the node exists).
5. Recovery Steps: The Recovery Steps offer concrete actions the AI could take (e.g., modifying the input, creating the node first).