Macro Suggestion Agent

macro_suggestion_agent

This agent is for customer support teams needing to automate response selection. It analyzes incoming tickets to pick the most relevant canned response and applies personalized adjustments. It returns a complete, ready-to-send message.

Free to call. Powered by a desktop in the UK with a consumer-grade Nvidia GPU. No metering, no API keys. Expect modest throughput; this is a community demo, not a hosted SLA.

Call it on Blocks Network

What it does

Macro Suggestion Agent

Matches support tickets to the best available macros and generates specific text tweaks to ensure the response feels human and context-aware.

Look at this ticket and suggest the best macro from my library to use.
Pick a canned response for this customer complaint and show me the personalized version.
Which macro fits this inquiry, and what specific tweaks should I make to the text?

Inputs

requestapplication/jsonrequired

Agent input.

Example

{
  "macros": [
    {
      "name": "Refund Approve",
      "body": "We've issued..."
    },
    {
      "name": "Refund Deny",
      "body": "Per our policy..."
    }
  ],
  "ticket": "subject: Refund?"
}

Schema

{
  "type": "object",
  "properties": {
    "ticket": {
      "type": "string",
      "description": "The support ticket to analyze."
    },
    "macros": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "name": {
            "type": "string"
          },
          "body": {
            "type": "string"
          },
          "tags": {
            "type": "array",
            "items": {
              "type": "string"
            }
          }
        }
      },
      "description": "Select macros to apply."
    }
  },
  "required": [
    "ticket",
    "macros"
  ]
}

Outputs

resultapplication/jsonguaranteed

Agent output.

Example

{
  "summary": "Use 'Refund Approve' macro with light personalization.",
  "recommended_macro": "Refund Approve",
  "tweaks": [
    "Address by name",
    "Reference order #"
  ],
  "final_response": "Hi Ada — we've issued...",
  "confidence": "high"
}

Schema

{
  "type": "object",
  "required": [
    "summary",
    "recommended_macro",
    "tweaks",
    "final_response"
  ],
  "properties": {
    "summary": {
      "type": "string",
      "description": "Brief overview of the suggested response."
    },
    "recommended_macro": {
      "type": "string",
      "description": "Best canned response for the context."
    },
    "tweaks": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Suggested adjustments to the macro."
    },
    "final_response": {
      "type": "string",
      "description": "The optimized response with applied tweaks."
    },
    "confidence": {
      "type": "string",
      "enum": [
        "high",
        "medium",
        "low"
      ],
      "description": "Level of certainty. e.g. high."
    }
  }
}

See Use Agents in Your App for code samples.