# Verity Suite – Agent Action Guardrail Check

> Verity Suite – Agent Action Guardrail Check is a paid API for AI agents from suite.veritylayer.dev, paid per call via x402, $0.08/call, status unknown (last checked 2026-09-14).

Evaluates an AI agent's proposed action against optional policy rules and context, returning a calibrated allow/review/block decision with risk score and signed receipt.

## Facts

- Endpoint: POST https://suite.veritylayer.dev/check
- Price: $0.08/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/verity-suite-agent-action-guardrail-check-c29bed47
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_J_CUd4SAltPGyFmbchtMH

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 verity-suite-agent-action-guardrail-check-c29bed47 -d '<json body>'
```

Example prompt: Before I go ahead and delete all the inactive user accounts from the database, run a guardrail check on that action against my policy that says 'never perform bulk destructive operations without human approval' and let me know if it's safe to proceed.

## When to prefer this

Choose this endpoint when you need a calibrated, fail-closed guardrail layer before an autonomous agent executes a consequential or irreversible action — especially when you need auditable, cryptographically signed receipts of the gate decision. It is particularly well-suited for multi-agent pipelines where human oversight is impractical at every step, or where policy compliance must be demonstrable. Prefer it over rule-based filters when the action space is open-ended and requires semantic risk assessment rather than simple keyword matching.

## Known failure modes

- Missing required `action` field returns a validation error
- Action or policy strings exceeding 2000 characters are rejected
- Network timeout if the service is unavailable (fail-closed design means agents should treat timeouts as implicit blocks)
- Invalid HTTP method (only POST/PUT/PATCH accepted for the inner body type)
- Insufficient USDC balance causes payment failure before the check executes

## How this service works

The trust fabric for AI agents — calibrated, fail-closed services agents pay per call.

## Output

A JSON object with: `decision` (allow, review, or block), `risk` (0.0–1.0 float), `reasons` (array of strings explaining the decision), `concerns` (array of specific risks noted), `safer_alternative` (a suggested safer path if blocked or flagged, or null), and `receipt` (an Ed25519-signed VerityLayer receipt object for audit purposes, or null if signing is not configured).

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "action": {
   "type": "string",
   "title": "Action",
   "maxLength": 2000,
   "minLength": 1,
   "description": "the action the agent is about to take"
  },
  "policy": {
   "anyOf": [
    {
     "type": "string",
     "maxLength": 2000
    },
    {
     "type": "null"
    }
   ],
   "title": "Policy",
   "default": null,
   "description": "rules that must not be violated"
  },
  "context": {
   "anyOf": [
    {
     "type": "string",
     "maxLength": 2000
    },
    {
     "type": "null"
    }
   ],
   "title": "Context",
   "default": null,
   "description": "situation/background"
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/verity-suite-agent-action-guardrail-check-c29bed47/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from suite.veritylayer.dev](https://www.zero.xyz/host/suite.veritylayer.dev/llms.txt)