Risk Summary Writer

Write a plain-language risk summary from the results of company or invoice checks, for a reviewer who is not technical.

Pay by wallet (x402) AI-generated risk-narrative-generate

Example

Example answer

{
  "flags": [
    {
      "finding": "The company domain is only 15 days old",
      "severity": "high",
      "recommendation": "Verify the company through additional channels before proceeding"
    }
  ],
  "summary": "1 of 3 checks flagged an issue. The domain age is under 30 days, which warrants attention.",
  "raw_checks": {
    "company_exists": {
      "pass": true
    },
    "domain_age_days": {
      "data": {
        "age_days": 15
      },
      "pass": false
    },
    "sanctions_clear": {
      "pass": true
    }
  },
  "risk_level": "low",
  "risk_score": 25,
  "passed_checks": [
    "company exists",
    "sanctions clear"
  ],
  "checks_performed": 3,
  "data_sources_consulted": [
    "Claude Sonnet risk assessment"
  ]
}

Inputs

Field	Type	Required	What it is
`context`	one of kyb, invoice_fraud	Yes	Context for the risk assessment (kyb or invoice_fraud)
`check_results`	object	Yes	Structured check results from upstream capabilities

Output fields

How reliably each field appears in a successful answer.

Field	Type	Appears
`flags`	array	Always (present in every successful answer)
`summary`	string	Always (present in every successful answer)
`raw_checks`	object	Always (present in every successful answer)
`risk_level`	string	Always (present in every successful answer)
`risk_score`	number	Always (present in every successful answer)
`passed_checks`	array	Always (present in every successful answer)
`checks_performed`	integer	Always (present in every successful answer)
`data_sources_consulted`	array	Usually (present for most inputs)

How to call it

The same tool through each route. MCP clients reach it through Strale’s server at https://api.strale.io/mcp; the call below is what the agent sends.

# Needs an API key. max_price_cents caps what this call may cost.
curl -X POST https://api.strale.io/v1/do \
  -H "Authorization: Bearer $STRALE_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "capability_slug": "risk-narrative-generate",
    "inputs": {
      "check_results": {
        "company_exists": {
          "pass": true
        }
      },
      "context": "kyb"
    },
    "max_price_cents": 5
  }'

{
  "tool": "strale_execute",
  "arguments": {
    "slug": "risk-narrative-generate",
    "inputs": {
      "check_results": {
        "company_exists": {
          "pass": true
        }
      },
      "context": "kyb"
    },
    "max_price_cents": 5
  }
}

import { Strale } from "straleio";

const strale = new Strale({ apiKey: process.env.STRALE_API_KEY! });
const result = await strale.do({
  capability_slug: "risk-narrative-generate",
  inputs: {
    "check_results": {
      "company_exists": {
        "pass": true
      }
    },
    "context": "kyb"
  },
  max_price_cents: 5,
});

import os
from straleio import Strale, DoRequest

strale = Strale(api_key=os.environ["STRALE_API_KEY"])
result = strale.do(DoRequest(
    capability_slug="risk-narrative-generate",
    inputs={
        "check_results": {
            "company_exists": {
                "pass": True,
            },
        },
        "context": "kyb",
    },
    max_price_cents=5,
))

# No account. The first request answers 402 with the price in USDC;
# an x402 client pays it from your wallet and repeats the request.
curl -i -X POST https://api.strale.io/x402/v2/risk-narrative-generate \
  -H 'Content-Type: application/json' \
  -d '{
    "check_results": {
      "company_exists": {
        "pass": true
      }
    },
    "context": "kyb"
  }'

Limitations

AI-generated narrative may oversimplify complex risk scenarios
Risk narratives are generated by AI and may oversimplify complex scenarios. The narrative should be reviewed by a qualified professional for high-stakes decisions.
Wording rules are enforced but not guaranteed
The AI is instructed to follow specific wording rules (e.g., never say "safe to pay") but adherence is not 100% guaranteed. Automated wording checks are recommended.
Context-dependent output quality
Narrative quality depends on the completeness of input check_results. Missing or incomplete check data will produce less useful narratives.