Prompt Injection Detection API
Catch prompt injections, role hijacks, and encoded jailbreaks in user inputs and RAG chunks.
Chain up to 5 Tinplate endpoints sequentially with automatic short-circuiting on security blocks. Interactive playground for POST /v1/pipeline: chain Prompt Injection Detection, PII Redaction, Secret Scanning, JSON Healing, and Command/SQL Guardrails in a single sub-millisecond edge request.
POST /v1/pipeline)
Ready
Select a preset security chain below (or customize the steps array) and click Run Live Tool to execute POST /v1/pipeline.
{
"ok": true,
"data": {
"verdict": "block",
"risk_score": 100,
"blocked": true,
"categories": [
"instruction_override",
"prompt_extraction"
],
"signals": [
{
"id": "instruction_override_ignore",
"category": "instruction_override",
"weight": 45,
"source": "normalized"
},
{
"id": "prompt_extraction_reveal",
"category": "prompt_extraction",
"weight": 45,
"source": "normalized"
}
],
"normalization": {
"stripped_invisible_chars": 0,
"folded_homoglyphs": 0,
"decoded_base64_segments": 0,
"decoded_other_segments": 0
},
"judge": "disabled",
"ai_judge": null
},
"meta": {
"request_id": "req_pg_01",
"ms": 1
}
}
This tool runs on the exact same stateless Worker engine as our production APIs. Unlock up to 256 KB payloads with a Polar All-Access key (/pricing):
Catch prompt injections, role hijacks, and encoded jailbreaks in user inputs and RAG chunks.
Mask emails, phones, IBANs, cards, and EU IDs with reversible tokens before LLM calls.