Batch Evaluate Policies

Evaluate up to 100 policy calls in a single HTTP request. Each item in the batch is an independent evaluation request. Results are returned in the same order as the input requests. Failures in individual items do not affect the rest of the batch.

Endpoint

POST /api/v1/policies/evaluate/batch

Required Role

MEMBER

Headers

Header	Value	Required
Authorization	Bearer <token>	Yes
Content-Type	application/json	Yes
X-Tenant-ID	Tenant identifier string	Yes

Request Body

{
  "requests": [
    {
      "policyModule": "string",
      "policyFunction": "string",
      "context": [{}]
    }
  ]
}

Field	Type	Required	Description
requests	array<EvaluationRequest>	Yes	Ordered list of evaluation requests. Maximum 100 items per call. Requests beyond index 99 are rejected with a 400 error.

EvaluationRequest

Each element of requests follows the same schema as the single /evaluate request body:

Field	Type	Required	Description
policyModule	string	Yes	Fully-qualified module name of the deployed policy.
policyFunction	string	Yes	Function name within the module to invoke.
context	array<object>	Yes	Ordered list of argument objects passed positionally to the function.

Response Body

{
  "responses": [
    {
      "result": "<any>",
      "executionTimeMs": 0,
      "error": null,
      "decisionTrace": null
    }
  ],
  "totalExecutionTimeMs": 0,
  "successCount": 0,
  "failureCount": 0
}

Field	Type	Description
responses	array<EvaluationResponse>	Results in the same order as the input requests array. Every request produces exactly one response entry, even if evaluation failed.
totalExecutionTimeMs	number	Combined wall-clock time in milliseconds for all evaluations plus batch coordination overhead.
successCount	number	Number of items where error is null.
failureCount	number	Number of items where error is non-null.

EvaluationResponse

Each element of responses has the same shape as the single /evaluate response:

Field	Type	Description
result	any	Return value of the policy function, or null if evaluation failed.
executionTimeMs	number	Execution time for this individual item.
error	string | null	Error message for this item; null on success.
decisionTrace	object | null	Always null — decision tracing is not available in batch mode.

HTTP Status Codes

Status	Meaning
200 OK	Batch processed. Inspect failureCount and per-item error fields to detect individual failures.
400 Bad Request	Malformed request body, requests array exceeds 100 items, or missing required fields.
401 Unauthorized	Missing or invalid bearer token.
403 Forbidden	Token is valid but the caller lacks the MEMBER role.
500 Internal Server Error	Unexpected engine failure affecting the entire batch.

Examples

curl

curl -X POST https://policy.aster-lang.dev/api/v1/policies/evaluate/batch \
  -H "Authorization: Bearer <token>" \
  -H "Content-Type: application/json" \
  -H "X-Tenant-ID: acme-corp" \
  -d '{
    "requests": [
      {
        "policyModule": "Loan.Approval",
        "policyFunction": "isEligible",
        "context": [
          { "applicantAge": 30, "creditScore": 720, "requestedAmount": 25000 }
        ]
      },
      {
        "policyModule": "Discount",
        "policyFunction": "calculate",
        "context": [
          { "order": { "total": 150 } }
        ]
      },
      {
        "policyModule": "Fraud.Detection",
        "policyFunction": "riskScore",
        "context": [
          { "transaction": { "amount": 9999, "country": "US" } }
        ]
      }
    ]
  }'

JavaScript

const requests = [
  {
    policyModule: 'Loan.Approval',
    policyFunction: 'isEligible',
    context: [{ applicantAge: 30, creditScore: 720, requestedAmount: 25000 }],
  },
  {
    policyModule: 'Discount',
    policyFunction: 'calculate',
    context: [{ order: { total: 150 } }],
  },
  {
    policyModule: 'Fraud.Detection',
    policyFunction: 'riskScore',
    context: [{ transaction: { amount: 9999, country: 'US' } }],
  },
];

const response = await fetch(
  'https://policy.aster-lang.dev/api/v1/policies/evaluate/batch',
  {
    method: 'POST',
    headers: {
      Authorization: 'Bearer <token>',
      'Content-Type': 'application/json',
      'X-Tenant-ID': 'acme-corp',
    },
    body: JSON.stringify({ requests }),
  }
);

const data = await response.json();

console.log(`Success: ${data.successCount}, Failure: ${data.failureCount}`);

data.responses.forEach((res, i) => {
  if (res.error) {
    console.error(`Request ${i} failed: ${res.error}`);
  } else {
    console.log(`Request ${i} result:`, res.result);
  }
});

Example Response

{
  "responses": [
    {
      "result": true,
      "executionTimeMs": 4,
      "error": null,
      "decisionTrace": null
    },
    {
      "result": 10,
      "executionTimeMs": 2,
      "error": null,
      "decisionTrace": null
    },
    {
      "result": null,
      "executionTimeMs": 1,
      "error": "Policy 'Fraud.Detection.riskScore' not found",
      "decisionTrace": null
    }
  ],
  "totalExecutionTimeMs": 9,
  "successCount": 2,
  "failureCount": 1
}