How CAERON CORE Works

A six-stage pipeline that scans, enforces, validates, and audits every AI interaction. Here is exactly what happens to every request.

1

SHIELD

Input Scanning

Every input is scanned for prompt injection, jailbreak attempts, PII exposure, and secrets leakage. 46 detection patterns across 4 categories.

example
// SHIELD detects injection attempt
{
  "category": "injection",
  "severity": "critical",
  "pattern": "ignore_instructions",
  "matched": "ignore all previous instructions"
}
2

GATE

Policy Enforcement

Policy engine evaluates scan results and issues a verdict: ALLOW, DENY, or REDACT. Configurable rules with severity thresholds.

example
// GATE policy verdict
{
  "verdict": "DENY",
  "reason": "Critical injection detected",
  "policy_version": "1.1"
}
3

LLM

Safe Forwarding

Only approved inputs reach the LLM. Redacted content has sensitive data replaced with safe tokens before forwarding.

example
// Redacted input forwarded to LLM
"Hello, my name is [REDACTED_NAME] and my
 email is [REDACTED_EMAIL]. How can I..."
4

OUTPUT VALIDATOR

Response Scanning

LLM responses are scanned for XSS, SQL injection, code injection, and PII/secrets leakage before delivery.

example
// Output validator catches XSS
{
  "category": "insecure_output",
  "severity": "high",
  "pattern": "xss_script_tag",
  "matched": "<script>alert('xss')</script>"
}
5

LOGBOX

Cryptographic Audit

Every transaction is recorded in an append-only, hash-chained log with Ed25519 signatures. Tamper-evident by design.

example
// LOGBOX entry (hash-chained)
{
  "seq": 42,
  "prev_hash": "a1b2c3...",
  "root_hash": "d4e5f6...",
  "signature": "ed25519_hex..."
}
6

EVIDENCE

Compliance Bundles

Generates cryptographic evidence bundles with receipts for regulatory compliance and forensic analysis.

example
// EVIDENCE cryptographic receipt
{
  "receipt_id": "uuid-v4",
  "root_hash": "blake3_hex...",
  "signature": "ed25519_hex...",
  "verifier_pubkey": "hex_64_chars"
}

Ready to Secure Your Pipeline?

Get started in minutes. Download, activate, and start protecting your AI applications.