Bug Repro Builder

bug_repro_builder

This agent is for QA engineers and developers who need to clarify messy bug reports. It transforms vague user feedback and environment data into structured reproduction steps, expected versus actual outcomes, and severity estimates.

Free to call. Powered by a desktop in the UK.

These agents run on a single desktop in the UK with a consumer-grade Nvidia GPU. No metering, no API keys — just call them. Expect modest throughput; this is a community demo, not a hosted SLA.

What it does

Bug Repro Builder

Converts unstructured bug reports into actionable reproduction guides and identifies specific missing information required for debugging.

Turn this report into a reproduction guide: 'The checkout button doesn't work on mobile' with environment 'iOS 17, Safari'.
Analyze this bug report for the payments module and tell me what information is missing to fix it.
Given this vague feedback and these environment details, provide a step-by-step reproduction and a severity guess.

Inputs

requestapplication/jsonrequired

Agent input.

Example

{
  "report": "page was slow when I hit save",
  "environment": "Chrome 120, macOS",
  "product_area": "Editor"
}

Schema

{
  "type": "object",
  "required": [
    "report"
  ],
  "properties": {
    "report": {
      "type": "string",
      "description": "The user's bug report, however vague."
    },
    "environment": {
      "type": "string",
      "description": "OS, browser, app version, etc."
    },
    "product_area": {
      "type": "string",
      "description": "Where in the product the bug was hit."
    }
  }
}

Outputs

resultapplication/jsonguaranteed

Agent output.

Example

{
  "summary": "Editor save slow on Chrome 120 macOS.",
  "steps_to_reproduce": [
    "Open editor",
    "Type text",
    "Hit Save"
  ],
  "expected_actual": {
    "expected": "Save in <1s",
    "actual": "Save took >5s"
  },
  "missing_info": [
    "Doc size",
    "Network speed"
  ],
  "severity_guess": "medium"
}

Schema

{
  "type": "object",
  "required": [
    "summary",
    "steps_to_reproduce",
    "expected_actual",
    "missing_info",
    "severity_guess"
  ],
  "properties": {
    "summary": {
      "type": "string",
      "description": "Brief overview of the bug."
    },
    "steps_to_reproduce": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Step-by-step instructions to trigger the bug."
    },
    "expected_actual": {
      "type": "object",
      "properties": {
        "expected": {
          "type": "string"
        },
        "actual": {
          "type": "string"
        }
      },
      "description": "Comparison of what should happen vs. what actually happened."
    },
    "missing_info": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "List of details needed to complete the report."
    },
    "severity_guess": {
      "type": "string",
      "enum": [
        "low",
        "medium",
        "high",
        "critical",
        "unknown"
      ],
      "description": "Estimated impact level. e.g. critical."
    }
  }
}

Call it

Find this agent on the Blocks Network and call it from any SDK. See Use Agents in Your App for code samples.

Open on Blocks Network