# Sentinel Agent Safety Guard

> Sentinel Agent Safety Guard is a paid API for AI agents from sentinel-agent.dev, paid per call via x402, $0.005/call, status unknown (last checked 2026-09-14).

Evaluates an agent action before execution and returns a pre-execution safety verdict to allow or block the action

## Facts

- Endpoint: POST https://sentinel-agent.dev/v1/guard
- Price: $0.005/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/sentinel-agent-safety-guard-dce8a8b5
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_Xbst-ZyVNEtGwHlLSCCif

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 sentinel-agent-safety-guard-dce8a8b5 -d '<json body>'
```

Example prompt: Before I let the agent call the 'transfer_funds' method of type 'financial_transaction', run a Sentinel safety check on that action and tell me whether it's safe to proceed.

## When to prefer this

Use this endpoint when an AI agent needs a safety gate before executing potentially dangerous, irreversible, or sensitive actions — especially financial transactions, system modifications, or external communications. Prefer Sentinel over post-execution logging when the cost of a bad action is high and you need a block/allow decision before the action runs.

## Known failure modes

- Missing required 'input.type' or 'input.method' fields returns a 400 validation error
- Malformed input object structure causes schema validation failure
- Network timeout if the safety evaluation takes too long
- Payment failure via x402 protocol results in 402 Payment Required before verdict is issued
- Unknown or unsupported action types may return an inconclusive verdict

## How this service works

SENTINEL pre-execution safety verdict

## Output

A pre-execution safety verdict indicating whether the specified agent action (identified by type and method) is allowed or should be blocked, along with any supporting safety assessment details.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "$schema": "https://json-schema.org/draft/2020-12/schema",
 "required": [
  "input"
 ],
 "properties": {
  "input": {
   "type": "object",
   "required": [
    "type",
    "method"
   ],
   "properties": {
    "type": {
     "type": "string"
    },
    "method": {
     "type": "string"
    }
   }
  },
  "output": {
   "type": "object",
   "required": [
    "type"
   ],
   "properties": {
    "type": {
     "type": "string"
    }
   }
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "risks": [],
  "verdict": "SAFE",
  "signature": "base64..."
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/sentinel-agent-safety-guard-dce8a8b5/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from sentinel-agent.dev](https://www.zero.xyz/host/sentinel-agent.dev/llms.txt)