I will test your llm app or ai agent for prompt injection and jailbreaks

A
ashok_dev_
A
ashok_dev_
Ashok Pasala

About this gig

If your product uses an LLM, users will try to make it ignore its instructions, leak its system prompt or data, or misuse its tools. I attack your chatbot or AI agent the way real users will, then show you exactly how to fix it.


What I test

  • Direct prompt injection and jailbreaks
  • Indirect prompt injection through documents, web pages, e-mails and tool results (RAG and agents)
  • System-prompt and sensitive-data leakage
  • Unsafe tool use, excessive agency and insecure output handling
  • Guardrail and content-filter bypasses


You get a clear report: every working attack with the exact prompt, how it works, severity, and a concrete fix (prompt hardening, input/output filtering, tool permissions, architecture changes). Premium includes a re-test after you patch.


Aligned with the OWASP Top 10 for LLM Applications. I only test systems you own or are authorised to test, and I am happy to sign an NDA.

Get to know Ashok Pasala

Ashok Pasala

Web and AI developer, websites, fixes, AI chatbots and agents

  • FromIndia
  • Member sinceJun 2026
  • Avg. response time1 hour
  • Languages

    English, Hindi
I run Norveth, a small web and AI studio. I build fast websites for small businesses, fix slow or broken sites, set up Google Business Profiles, and build AI chatbots, agents and automations. Stack: Next.js, React, WordPress, Shopify; Python, LangChain, OpenAI, Claude, Gemini; n8n, Make, Zapier. I also test AI apps for prompt injection and data leaks. Fixed price and a clear plan before I start, 1 to 3 day delivery for small jobs, and nothing changes on your live site until you approve it. Message me with your site or idea for a plan within a few hours.