# Agent Security Gateway – Schema Drift Check

> Agent Security Gateway – Schema Drift Check is a paid API for AI agents from agent-security-gateway.onrender.com, paid per call via x402, $0.01/call, status unknown (last checked 2026-09-15).

Detects dangerous or unexpected field additions in API response content to prevent schema drift attacks against AI agents

## Facts

- Endpoint: POST https://agent-security-gateway.onrender.com/api/schema/drift-check
- Price: $0.01/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-15
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/agent-security-gateway-schema-drift-check-3b7514ce
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_deScf4J2r91K5C6bsFAls

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability agent-security-gateway-schema-drift-check-3b7514ce -d '<json body>'
```

Example prompt: Before I process this tool response, run a schema drift check on it — the response content is '{"user":"alice","admin_token":"sk-xxx"}' from the 'user-lookup' tool (agent ID: agent-42) — flag any dangerous new fields that weren't expected.

## When to prefer this

Use this endpoint when an AI agent receives a response from an external tool or API and needs to verify that no unexpected or dangerous fields (like auth tokens, admin keys, or privilege escalation vectors) have been added before acting on the response. Particularly valuable in autonomous agent pipelines where tool outputs are trusted and schema drift could be exploited as an attack vector.

## Known failure modes

- Missing required 'response_content' field returns a 400 validation error
- Malformed JSON in response_content may cause parsing failures
- Service unavailability on render.com free tier may cause timeouts or cold-start delays
- Payment failure via x402/USDC may prevent request from being processed
- Empty response_content may produce low-confidence or trivially 'allow' results

## How this service works

Pay-per-request security APIs for autonomous AI agents using x402. Detect Japanese prompt injection, validate content, scan for threats, and perform pre-payment security checks. Built for USDC/Base payments.

## Output

Returns a JSON object with an 'allow' boolean, a 'decision' field ('allow' or 'block'), a 'risk_level' ('high', 'medium', 'low'), a list of 'reasons' explaining what was flagged (e.g. 'dangerous_new_fields: [admin_token]'), and a 'recommended_action' (e.g. 'reject_schema_update'). The 'primitive' field identifies this as a 'schema-drift-check' operation.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "allow": {
   "type": "boolean"
  },
  "reasons": {
   "type": "array"
  },
  "decision": {
   "type": "string"
  },
  "primitive": {
   "type": "string"
  },
  "risk_level": {
   "type": "string"
  },
  "recommended_action": {
   "type": "string"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "allow": false,
  "reasons": [
   "dangerous_new_fields: ['admin_token']"
  ],
  "decision": "block",
  "primitive": "schema-drift-check",
  "risk_level": "high",
  "recommended_action": "reject_schema_update"
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agent-security-gateway-schema-drift-check-3b7514ce/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agent-security-gateway.onrender.com](https://www.zero.xyz/host/agent-security-gateway.onrender.com/llms.txt)