I will perform ai, llm and agent security penetration testing

C
cyber_cypher
C
cyber_cypher
Rehan M

Level 1

About this gig

Looking for a hands-on security assessment of your AI application, LLM chatbot, RAG system, or AI agent?

I provide manual AI/LLM penetration testing and red teaming focused on identifying real attack paths beyond automated checks. Testing can include:

  • Direct & indirect prompt injection
  • Jailbreaks and system prompt leakage
  • Sensitive data disclosure
  • Unsafe output handling
  • RAG retrieval and knowledge-base manipulation
  • Authentication, authorization & API weaknesses
  • Excessive agency and unsafe tool/function usage
  • Agent workflow manipulation
  • Business logic and chained attack scenarios
  • Supporting web/API security issues


Assessments can be performed in black-box or grey-box mode and aligned with recognized OWASP guidance for LLM and agentic application security.


You receive a professional report with validated findings, severity, evidence, impact, attack scenarios, and practical remediation guidance. Remediation retesting can also be provided.

I can tailor testing for multiple models, agents, RAG pipelines, tools, plugins, and custom AI architectures.

Get to know Rehan M

Rehan M

Offsec rookie

4.9(9)

Level 1

  • FromPakistan
  • Member sinceMar 2023
  • Avg. response time1 hour
  • Last delivery1 week
  • Languages

    English, German, Arabic, Turkish
Hi, I'm Rehan, a cybersecurity professional specializing in penetration testing and application security. I assess web applications, APIs, networks, and Active Directory environments using a hands-on, attacker-focused approach that goes beyond automated scans. My work includes vulnerability assessment, manual validation, business logic testing, reconnaissance, and clear remediation guidance. I hold OSCP, CRTP, eCPPTv2, and Certified AppSec Practitioner certifications. Need a security assessment or custom scope? Feel free to message me.