# AI Security Guard Advisory Scan

> AI Security Guard Advisory Scan is a paid API for AI agents from aisecurityguard.io, paid per call via x402, $0.0195/call, status unknown (last checked 2026-09-14).

Scans external content (API responses, URLs, documents) for prompt injection, malicious code, and policy violations before an AI agent processes it

## Facts

- Endpoint: POST https://aisecurityguard.io/v1/advisory
- Price: $0.0195/call
- Payment: x402
- Status: unknown
- Last checked: 2026-09-14
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/ai-security-guard-1219e44a
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_8GfuB4mVVhgorkmvdMhcg

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 ai-security-guard-1219e44a -d '<json body>'
```

Example prompt: Before you process the response from that external MCP tool, run it through AI Security Guard first — scan the content for prompt injection, malicious code, or anything trying to hijack your behavior, and tell me if it's safe to proceed.

## When to prefer this

Use this endpoint when your AI agent needs to process content from untrusted external sources — MCP tool responses, fetched URLs, third-party API payloads, or user-supplied documents — and you need a deterministic, non-LLM security layer to detect prompt injection or malicious instructions before the agent acts on them. Prefer this over LLM-based self-checking (which is itself vulnerable to injection) or no validation at all.

## Known failure modes

- Malformed or oversized payload returns 400 Bad Request
- Authentication or payment verification failure returns 402 or 401
- Upstream analyzer timeout returns 503 with partial results
- Content encoding issues may cause misclassification
- False positives on benign but unusual-looking content

## How this service works

When your AI agent calls an MCP tool, fetches a URL, or processes a document, it trusts whatever comes back. That's the problem. A malicious API response can hijack your agent's behavior or tell it to run malicious code. AI Security Guard is a firewall that sits between your agent and untrusted content. Before your agent processes anything external, we scan it. Five specialized analyzers — not LLMs, but deterministic detection systems — examine inputs and provide verdicts with context. When something looks wrong, agents get an explanation of what it might be and why it matters. Operators get visibility into what's flowing through their systems. x402 micropayments. Pay per scan, no subscriptions.

## Output

Returns a security advisory with verdicts from five deterministic analyzers, including whether the content is safe, what type of threat was detected (if any), an explanation of why it matters, and context to help the agent or operator decide whether to proceed.

## Example request

```json
{
 "context": "web_agent using external_apis and user_input data sources",
 "question": "How can I detect and prevent prompt injection attacks in my AI agent that processes user-submitted content from external APIs?"
}
```

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "context": {
   "type": "object",
   "description": "Optional context about your agent: agent_type (mcp_client, web_agent), data_sources (external_apis, user_input, file_system), etc."
  },
  "question": {
   "type": "string",
   "description": "Your security question about AI agent threats, defenses, or best practices. Can be general or specific to your use case."
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "advisory": "MCP servers with file system and database access require layered protection: 1) Validate all tool inputs against allowlists, 2) Implement path traversal prevention for file operations, 3) Use parameterized queries for database access, 4) Scan all external data before processing.",
  "confidence": 0.91,
  "recommendations": [
   "Implement input sanitization for all tool parameters",
   "Use AI Security Guard to scan web_fetch responses before processing",
   "Create audit logs for all file system operations"
  ],
  "threat_categories": [
   "tool_abuse",
   "data_exfiltration",
   "path_traversal"
  ]
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/ai-security-guard-1219e44a/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from aisecurityguard.io](https://www.zero.xyz/host/aisecurityguard.io/llms.txt)