AI agent red teaming platform
Prove it's safe. Govern every call — in your cluster.
MrProbe, by QDEX Consulting (Chennai, India), is an enterprise AI agent assurance platform for red teaming and functional evaluation.
MrProbe is an AI agent red teaming platform, an LLM red teaming tool, and the product teams use for AI agent security testing. It is built to prove an agent is safe, and to govern every call in your cluster.
The security catalogue is 648+ techniques across 17 categories. The category names are Prompt Injection, Jailbreaking, Data Exfiltration, Agent Specific, Multi-Agent, Content Safety, PII Hygiene & Reflection, Ground-Truth PII Extraction, RAG Exploitation, Output-to-Downstream, Multilingual, Functional Quality, Adversarial ML, Bias & Fairness, Email Agent, Training Data Poisoning, Supply Chain. A live campaign can add jurisdiction-specific probes on top of that list.
Functional evaluation uses 20 dimensions: Functional, Groundedness, Hallucination, Factual Precision, Completeness, Coherence & Fluency, Conciseness, Instruction Adherence, Deviation Validation, Task Completion, Business Alignment, Toxicity, PII Leakage, Bias & Fairness, Regulatory Compliance, Consistency, Robustness, Context Retention, Multi-turn Coherence, Citation Accuracy. A run can be repeated against a golden set so a later change shows up as a regression, not as a new anecdote.
Country-gated PII covers 16 jurisdictions: United States, European Union, United Kingdom, India, Japan, China, Singapore, United Arab Emirates, Canada, Australia, Brazil, South Korea, Saudi Arabia, Indonesia, South Africa, Türkiye. Findings map to 30 named frameworks (13 global and 17 country-specific): OWASP LLM Top 10, NIST AI RMF, MITRE ATLAS, ISO/IEC 42001, OWASP Top 10 for Agentic Applications, CSA Confused Deputy, GDPR, UK GDPR, CCPA / CPRA, HIPAA, India DPDP, Japan APPI, PIPEDA · Law 25, Singapore PDPA, Brazil LGPD, Korea PIPA, KSA PDPL, PCI DSS v4, SOX, GLBA, FedRAMP, FFIEC, EU AI Act, NIS2, DORA, UAE PDPL, DIFC DPL, ADGM DPR, MAS TRM, BACEN. The two agentic names in that list are OWASP Top 10 for Agentic Applications and CSA Confused Deputy, both registered in the product. MrProbe does not claim ISO 42001 certification.
LENS is the testing product. ATC is the control path. On the self-hosted ATC path, prompts, logs, and the ledger stay in the customer cluster. The hosted app at mrprobe.ai is where a team registers an agent, starts a campaign, and reads the report.
A standard assessment is a campaign against one registered agent. Security findings carry a verdict. Functional scores sit in the same report. CI can call the API and fail a build when the gate is not met. That gate is a product surface, not a sidecar script the customer has to invent.
What you can open next
Red teaming, evaluation, compliance, deployment, and pricing each have their own URL. Documentation is published on mrprobe.ai/docs, including getting started and compliance mapping.
Questions
How is MrProbe different from a fixed prompt list?
MrProbe probes the agent surface you register and attaches evidence to each finding. A result is marked proven, suspected, or defended. The product does not ask a team to treat an unmatched prompt hit as a confirmed exploit.
Does MrProbe include functional evaluation and not only security red teaming?
Yes. The same product runs adversarial testing and a 20-dimension functional evaluation. Quality, grounding, consistency, and regression-ready reruns sit next to the security campaign.
What compliance frameworks does MrProbe map?
MrProbe maps findings to 30 named frameworks (13 global and 17 country-specific): OWASP LLM Top 10, NIST AI RMF, MITRE ATLAS, ISO/IEC 42001, OWASP Top 10 for Agentic Applications, CSA Confused Deputy, GDPR, UK GDPR, CCPA / CPRA, HIPAA, India DPDP, Japan APPI, PIPEDA · Law 25, Singapore PDPA, Brazil LGPD, Korea PIPA, KSA PDPL, PCI DSS v4, SOX, GLBA, FedRAMP, FFIEC, EU AI Act, NIS2, DORA, UAE PDPL, DIFC DPL, ADGM DPR, MAS TRM, BACEN. Mapping a control is not a certification. MrProbe does not claim ISO 42001 certification.
Published by QDEX Consulting.