Whether the risk model plane can actually work right now

GET /v1/risk/health

Reports whether the per-organisation model plane is genuinely usable: that the plane was built, that the per-organisation stores can be written, and whether the event surface the feature plane is rolled up from is reachable. It is a REAL probe, not status theatre.

200 only when the plane can work. Otherwise 503 CARRYING THE REPORT — which part failed and the real error — and that body is why this is not a typed op: a typed op reaches a non-2xx by returning an error, and the envelope that produces would drop exactly the detail the probe exists to deliver.

An unreachable event surface is REPORTED and is not a failure. Scoring reads in-memory aggregates and never the warehouse, so a warm that cannot run degrades how much history a model has seen and does not stop it deciding.

It also reports how many organisations' models are resident, how many have been evicted to hold that bound, and how many of the resident ones are at their own aggregate bound. Eviction is lossless — learned state is written to that organisation's own store first and its aggregates rebuild from its own record — so a climbing count is a capacity signal, not a loss. A STRAINED model is different: it has started forgetting its own least-recently-active subjects, and each forgotten subject reads as inactive until it is active again. That is a control degrading, and it is reported here because it is otherwise silent.

It answers about the process, not about a tenant: it takes no organisation and names none.

Request

GET /v1/risk/health takes no parameters and no body — the credential is the whole request.

Response

The document declares no response body for this operation. It answers 200 on success and the platform error shape on failure — see Errors.

Examples

hanzo risk health

import { Configuration, RiskApi } from 'hanzoai';

const api = new RiskApi(new Configuration({ accessToken: process.env.HANZO_API_KEY }));
const { data } = await api.getRiskHealth();

from hanzoai.cloud import ApiClient, Configuration
from hanzoai.cloud.api import RiskApi

client = ApiClient(Configuration(access_token=os.environ["HANZO_API_KEY"]))
result = RiskApi(client).get_risk_health()

cfg := cloud.NewConfiguration()
cfg.AddDefaultHeader("Authorization", "Bearer "+os.Getenv("HANZO_API_KEY"))
client := cloud.NewAPIClient(cfg)

resp, _, err := client.RiskAPI.GetRiskHealth(context.Background()).Execute()
if err != nil {
	return err
}

use hanzo_cloud::apis::{configuration::Configuration, risk_api};

let mut cfg = Configuration::new();
cfg.bearer_access_token = std::env::var("HANZO_API_KEY").ok();

let result = risk_api::get_risk_health(&cfg, Default::default()).await?;

import ai.hanzo.cloud.ApiClient;
import ai.hanzo.cloud.api.RiskApi;

ApiClient client = new ApiClient();
client.setRequestInterceptor(b -> b.header("Authorization", "Bearer " + System.getenv("HANZO_API_KEY")));

var result = new RiskApi(client).getRiskHealth();

curl https://api.hanzo.ai/v1/risk/health \
  -H "Authorization: Bearer $HANZO_API_KEY"

curl -X POST https://api.hanzo.ai/v1/mcp \
  -H "Authorization: Bearer $HANZO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
       "jsonrpc": "2.0",
       "id": 1,
       "method": "tools/call",
       "params": {
         "name": "risk",
         "arguments": {
           "op": "get_risk_health",
           "input": {}
         }
       }
     }'

Risk API · All Hanzo APIs · Interactive reference