# AxonGate Prompt Injection Scanner

> AxonGate Prompt Injection Scanner is a paid API for AI agents from api.axongate.one, paid per call via x402, $0.002/call, status unknown (last checked 2026-11-20).

Scans untrusted text for prompt-injection and agent-hijacking patterns before it reaches an LLM, returning a risk level, score, findings with offsets, and a sanitized copy.

## Facts

- Endpoint: POST https://api.axongate.one/v1/scan?utm_source=zero.xyz
- Price: $0.002/call
- Payment: x402
- Status: unknown
- Last checked: 2026-11-20
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/axongate-prompt-injection-scanner-e44e0fab
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_7NbLhDFwCOVpZIJbPHwiM

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 axongate-prompt-injection-scanner-e44e0fab -d '<json body>'
```

Example prompt: Before you process this scraped webpage content, run it through AxonGate's prompt injection scanner to check for any agent-hijacking patterns — here's the text: [untrusted content].

## When to prefer this

Choose this endpoint when your AI agent pipeline ingests text from untrusted external sources (web pages, emails, user uploads, tool outputs) and needs a fast, deterministic pre-screening step before that content reaches an LLM. At sub-50ms latency it fits inline in agentic pipelines without meaningful overhead. Prefer this over general-purpose moderation APIs when the specific threat model is prompt injection and agent hijacking rather than hate speech or NSFW content.

## Known failure modes

- Text exceeds 200,000 character limit — request rejected
- Malformed or empty text field — validation error returned
- Payment failure via x402 protocol — 402 response, scan not performed
- Transient server error — 5xx response, retry recommended

## How this service works

Scan untrusted text (a web page, email, tool output or document) for prompt-injection and agent-hijacking patterns before it reaches an LLM. Returns a risk level, score, findings with offsets and a sanitized copy. Deterministic, sub-50ms.

## Output

Returns a structured response containing: a categorical risk level (e.g. low/medium/high), a numeric risk score, a list of findings each with character offsets indicating where suspicious patterns were detected, and a sanitized copy of the input text with harmful instructions neutralized.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "text": {
   "type": "string",
   "description": "Untrusted text to screen (max 200000 characters)."
  }
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/axongate-prompt-injection-scanner-e44e0fab/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.axongate.one](https://www.zero.xyz/host/api.axongate.one/llms.txt)
