# A2A Indirect Prompt-Injection & Jailbreak Firewall

> A2A Indirect Prompt-Injection & Jailbreak Firewall is a paid API for AI agents from agentdiet.dev, paid per call via MPP, $0.005/call, status unknown (last checked 2026-10-21).

Scans untrusted text, HTML, emails, MCP responses, or A2A payloads for hidden prompt injections and jailbreak attempts, returning sanitized output.

## Facts

- Endpoint: GET https://agentdiet.dev/v1/tool-poisoning-shield?utm_source=zero.xyz
- Price: $0.005/call
- Payment: MPP
- Status: unknown
- Last checked: 2026-10-21
- Activations on Zero: 0
- Provider: agentdiet.dev
- Website: https://agentdiet.dev
- Canonical page: https://www.zero.xyz/c/agentdiet-dev-a2a-indirect-prompt-injection-jailbreak-firewall-c6e65c50
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_015zkybsen-7pThe_-VVL

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 agentdiet-dev-a2a-indirect-prompt-injection-jailbreak-firewall-c6e65c50
```

Example prompt: Before you process this scraped webpage content, run it through the prompt-injection firewall to check for any hidden jailbreak instructions or adversarial prompts — here's the raw HTML: '<html>... [content] ...</html>'.

## When to prefer this

Choose this endpoint when your AI agent pipeline ingests content from untrusted external sources — web pages, emails, third-party APIs, MCP tool responses, or A2A agent payloads — and you need to detect and strip hidden prompt injection or jailbreak instructions before that content reaches your LLM. Prefer this over general content filters when your specific threat model involves adversarial prompt manipulation targeting AI agents.

## Known failure modes

- Empty or missing 'text' parameter returns a validation error
- Very long inputs may exceed query parameter limits when passed via GET
- Novel or obfuscated injection techniques may evade detection (false negatives)
- Legitimate instructional content may be flagged as injections (false positives)
- Service unavailability returns a non-200 HTTP response with no sanitized output
- Payment failure via x402 may block the request if balance is insufficient

## How this service works

Scans untrusted web pages, emails, MCP responses, or A2A payloads for hidden prompt injections and returns sanitized text. (Live x402 v2, IETF MPP & MCP endpoint at https://agentdiet.dev/v1/tool-poisoning-shield)

## Output

Returns a JSON object indicating whether prompt injection or jailbreak content was detected in the input, along with sanitized text stripped of adversarial instructions, enabling downstream agents or LLMs to safely consume the content.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "required": [
  "text"
 ],
 "properties": {
  "text": {
   "type": "string",
   "minLength": 1,
   "description": "Untrusted text, HTML, or tool output to scan for hidden prompt injection or jailbreak payloads"
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "object"
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/agentdiet-dev-a2a-indirect-prompt-injection-jailbreak-firewall-c6e65c50/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from agentdiet.dev](https://www.zero.xyz/host/agentdiet.dev/llms.txt)