# Security Injection Resilience Check

> Security Injection Resilience Check is a paid API for AI agents from api.craigmbrown.com, paid per call via x402, $0.5/call, status unknown (last checked 2026-09-14).

Tests whether an agent's input handling resists prompt-injection and content-trap attack patterns, returning structured findings and a trap scan result.

## Facts

- Endpoint: POST https://api.craigmbrown.com/v1/services/security.injection-resilience
- Price: $0.5/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/security-injection-resilience-check-7912cb16
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_PZRoUr38g1eYMy-bnK0Hs

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 security-injection-resilience-check-7912cb16 -d '<json body>'
```

Example prompt: Run a prompt-injection resilience check on my agent at 'agent-xyz' — I want to see whether it resists content traps and injection patterns, and get back the full findings report.

## When to prefer this

Choose this endpoint when you need a structured, verifiable resilience assessment of a specific agent's input handling — particularly before granting it spend authority, delegating sensitive tasks, or integrating an untrusted third-party agent. It provides a tamper-evident proof token (BlindOracle), making it more trustworthy than an ad-hoc manual review and better suited for agentic pipelines requiring audit trails.

## Known failure modes

- Missing or invalid 'target' field returns a 400-level error
- Target agent unreachable or unresponsive causes scan timeout
- Ambiguous target identifier leads to empty findings with no actionable output
- Rate limiting or payment failure blocks the scan from executing

## How this service works

Tests whether a counterparty agent's input handling resists prompt-injection and content-trap patterns — a concrete resilience check, not a compliance checkbox.

## Output

A JSON object containing a 'findings' array listing detected vulnerabilities or passed checks, a 'trap_scan' string summarizing content-trap scan results, and a 'proof_kind' integer (30105) serving as a cryptographic proof token confirming the deliverable was produced by BlindOracle.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "target": {
   "type": "string"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "schema": {
  "type": "object",
  "properties": {
   "findings": {
    "type": "array"
   },
   "trap_scan": {
    "type": "string"
   },
   "proof_kind": {
    "type": "integer",
    "const": 30105
   }
  }
 },
 "example": {
  "trap_scan": "..."
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/security-injection-resilience-check-7912cb16/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.craigmbrown.com](https://www.zero.xyz/host/api.craigmbrown.com/llms.txt)