Policy Consistency Checker

policy_consistency_checkerSpecialist

This agent is for customer experience teams to audit draft communications. It compares proposed responses against official company policies to identify contradictions and returns a summary of discrepancies with suggested corrections.

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

Policy Consistency Checker

Detects direct mismatches between draft text and policy documentation to ensure compliance without interpreting intent.

Check this draft email against our refund policy and list any conflicts.
Does this support response contradict the published shipping guidelines?
Review this message for policy consistency and suggest a fix for any errors.

Inputs

requestapplication/jsonRequired

Agent input.

Example

{
  "response": "We'll refund any time within 30 days.",
  "policy": "Refunds within 14 days, items unused."
}

Schema

{
  "type": "object",
  "required": [
    "response",
    "policy"
  ],
  "properties": {
    "response": {
      "type": "string",
      "description": "The content to check against the policy."
    },
    "policy": {
      "type": "string",
      "description": "The policy to check against."
    }
  }
}

Outputs

resultapplication/jsonGuaranteed

Agent output.

Example

{
  "summary": "1 conflict on refund window.",
  "consistent": false,
  "conflicts": [
    {
      "issue": "Refund window mismatch",
      "policy_excerpt": "14 days",
      "response_excerpt": "30 days",
      "severity": "high"
    }
  ],
  "suggested_correction": "Per our policy refunds are within 14 days..."
}

Schema

{
  "type": "object",
  "required": [
    "summary",
    "consistent",
    "conflicts"
  ],
  "properties": {
    "summary": {
      "type": "string",
      "description": "Brief overview of policy discrepancies found."
    },
    "consistent": {
      "type": "boolean",
      "description": "Whether the response aligns with published policy."
    },
    "conflicts": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "issue": {
            "type": "string"
          },
          "policy_excerpt": {
            "type": "string"
          },
          "response_excerpt": {
            "type": "string"
          },
          "severity": {
            "type": "string",
            "enum": [
              "low",
              "medium",
              "high"
            ]
          }
        }
      },
      "description": "List of detected policy contradictions."
    },
    "suggested_correction": {
      "type": "string",
      "description": "Proposed fix for the policy violation."
    }
  }
}

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