A2A Indirect Prompt-Injection & Jailbreak Firewall

Choose this endpoint when your AI agent is ingesting content from untrusted sources — scraped web pages, third-party tool responses, MCP payloads, emails, or user-submitted text — and you need to screen for hidden prompt-injection or jailbreak payloads before passing content to an LLM. Prefer it over general-purpose content filters when the threat model is specifically adversarial LLM instruction injection rather than spam, toxicity, or malware.

• Empty or missing 'text' parameter returns a 400 validation error
• Very long payloads may exceed query parameter length limits for GET requests
• False negatives: sophisticated obfuscated injections may not be detected
• False positives: legitimate instructional text may be flagged as injection
• Payment required (HTTP 402) if x402 micropayment not included
• Malformed HTML may produce incomplete sanitization