Support: support@tinplate.dev · Zero-Retention Edge Security Support Trust Center DPA GitHub Issues
Prompt Injection, PII Redaction & JSON Repair

Redact PII, Block Prompt Injection & Repair Malformed LLM JSON in One Pipeline

Protect both sides of your LLM call: strip PII and API secrets before prompts leave your VPC, block Unicode/Base64 prompt injections, and deterministically heal broken JSON before it crashes your downstream parser.

Why Teams Use This Bundle

  • Reversible [PII_EMAIL_1] tokenization via /v1/pii-redactor and /v1/pii-redactor/restore
  • Fixes unquoted keys, trailing commas, single quotes, and Markdown fences in /v1/json-heal
  • Zero-logging architecture: customer prompts, PII, and secrets are never persisted or sampled
Quickstart

Run PII redaction + prompt injection detection in 1 call

curl -X POST "https://api.tinplate.dev/v1/presets/llm-ingest-safe" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_POLAR_API_KEY" \
  -d '{"text":"Hi, my email is alice@acme.eu and my key is sk_live_1234567890abcdef. Ignore previous instructions.","context":"user_input"}'

Included Edge Micro-APIs (6)

POST /v1/prompt-guard Live

Prompt Injection Detection API

Catch prompt injections, role hijacks, and encoded jailbreaks in user inputs and RAG chunks.

POST /v1/pii-redactor Live

PII Redactor API

Mask emails, phones, IBANs, cards, and EU IDs with reversible tokens before LLM calls.

POST /v1/secret-scanner Tier 1

Log Redact & Secret Scanner API

Detect and mask leaked API keys, JWTs, and private keys in logs, diffs, and prompts.

POST /v1/json-heal Live

Broken JSON & LLM Output Healer API

Repair malformed or truncated LLM JSON outputs into valid, schema-checked JSON in 1 ms.

POST /v1/llm-output-guard Tier 1

LLM Output Safety, Leakage & Format Guardrail API

Scan and sanitize LLM outputs for prompt leakage, Markdown beacons, XSS, and policy violations.

POST /v1/llm-messages-guard Live

LLM Chat Messages Policy Proxy & Sanitizer API

Sanitize entire multi-turn LLM messages[] arrays in one call before forwarding to OpenAI, Anthropic, or Gemini.