# neo injection-scan

> neo injection-scan is a paid API for AI agents from neo-collettive.onrender.com, paid per call via x402, $0.01/call, status unknown (last checked 2026-10-29).

Scans text or a public URL for prompt injection attacks including overrides, fake roles, exfiltration attempts, tool abuse, payment requests, hidden Unicode/HTML instructions, and base64-encoded payloads, returning signed rule-based findings.

## Facts

- Endpoint: POST https://neo-collettive.onrender.com/x402/injection-scan?utm_source=zero.xyz
- Price: $0.01/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-29
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/neo-injection-scan-35a81e42
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_N7NUta95LzrT_6jMtG7ZT

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 neo-injection-scan-35a81e42 -d '<json body>'
```

Example prompt: Before you process this page at https://suspicious-site.com/instructions.html, scan it for prompt injection — I want to know if it's trying to override your role, steal data, or sneak in payment requests.

## When to prefer this

Choose this endpoint when your agent needs to safely ingest untrusted external content — scraped pages, user-submitted text, third-party documents — before passing it to an LLM or tool-calling pipeline. It is specifically designed for agentic security contexts where prompt injection is a real threat, providing signed, rule-based evidence (not just a probability score) so you have an auditable record. Prefer it over generic content moderation APIs when the threat model specifically involves instruction injection, role overrides, hidden Unicode, or payment manipulation targeting AI agents.

## Known failure modes

- URL is not publicly accessible or returns non-200 — scan cannot fetch page content
- Text exceeds 200,000 character limit — request rejected
- Invalid JSON body or missing both url and text fields — bad request error
- Render cold-start latency causing timeout on first call
- Signature verification fails if log integrity is compromised

## How this service works

Scan text or a public page for prompt injection before your agent reads it: overrides, fake roles, exfiltration, tool abuse, payment asks, invisible Unicode/HTML-hidden instructions, base64. Rule-based evidence, signed and logged.

## Output

A JSON object with a risk level (e.g. 'high', 'medium', 'low'), an array of findings each containing the type of injection (e.g. 'override', 'exfiltration', 'tool_abuse') and supporting evidence text, a numeric risk_score (0-100), and a cryptographic signature confirming the scan result is authentic and logged.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "url": {
   "type": "string",
   "description": "public https page, instead of text"
  },
  "text": {
   "type": "string",
   "maxLength": 200000
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "level": "high",
  "findings": [
   {
    "type": "override",
    "evidence": "..."
   }
  ],
  "signature": "...",
  "risk_score": 70
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/neo-injection-scan-35a81e42/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from neo-collettive.onrender.com](https://www.zero.xyz/host/neo-collettive.onrender.com/llms.txt)