LLM Security Audit — Prompt Injection & Jailbreak Assessment

I will conduct a security assessment of your LLM application — testing for prompt injection vulnerabilities, jailbreak susceptibility, data exfiltration through adversarial prompts, system prompt extraction, and insecure output handling — and delivering a security report with severity-rated findings and specific remediation recommendations for every vulnerability discovered. LLM applications face a category of security vulnerability that conventional application security tools cannot detect — prompt injection attacks that hijack the LLM's behaviour through malicious user input are the OWASP Top 10 for LLM Applications number-one risk, and most LLM applications deployed without security testing are vulnerable to at least basic injection attacks.

The assessment covers direct and indirect prompt injection testing, system prompt extraction attempts, jailbreak testing across common techniques, data exfiltration via adversarial prompts, insecure output handling review, and a security report with CVSS-equivalent severity ratings and specific remediation code or configuration for each finding.