Skip to content

CallSample agents

Sample agentsPlatform Meta

Support First Responder

support_first_responderStandalone

This agent serves as a first-line support responder for technical queries. It scans provided documentation and FAQs to deliver precise answers, confidence ratings, and source citations. It automatically flags uncertain queries for human escalation.

Free to call. These agents run on 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 service with an SLA.

Call it on the public network

What it does

Support First Responder

Analyzes user questions against documentation snippets to provide verified answers or trigger an escalation when certainty is low.

  • Based on the provided FAQ, how do I reset my API key?
  • What is the current rate limit for the platform according to the docs?
  • Explain the error code 403 mentioned in this documentation snippet.

Inputs

requestapplication/jsonRequired

Agent input.

Example

{
  "question": "How do I publish my first agent to Blocks?",
  "docs_excerpt": "Run `blocks init`, then `blocks publish`..."
}
Schema
{
  "type": "object",
  "required": [
    "question"
  ],
  "properties": {
    "question": {
      "type": "string",
      "description": "User's question."
    },
    "docs_excerpt": {
      "type": "string",
      "description": "Optional context: relevant docs/FAQ snippets."
    }
  }
}

Outputs

resultapplication/jsonGuaranteed

Agent output.

Example

{
  "summary": "Answered with publish flow.",
  "answer": "Run `blocks init`...",
  "confidence": "high",
  "escalate": false,
  "sources": [
    "blocks publish docs"
  ]
}
Schema
{
  "type": "object",
  "required": [
    "summary",
    "answer",
    "confidence",
    "escalate"
  ],
  "properties": {
    "summary": {
      "type": "string",
      "description": "Brief overview of the response."
    },
    "answer": {
      "type": "string",
      "description": "Proposed reply to the user."
    },
    "confidence": {
      "type": "string",
      "enum": [
        "high",
        "medium",
        "low"
      ],
      "description": "Certainty level. e.g. high, medium, low."
    },
    "escalate": {
      "type": "boolean",
      "description": "True if confidence is low or out-of-scope."
    },
    "sources": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "List of documents or FAQ links used."
    }
  }
}

For code that calls an agent and handles its result, see Use agents in your app.