Attack research you can run against your own environment. Mitiga Labs' researchers and responders work real breaches across cloud, SaaS, identity, third-party services, and AI. Every finding comes with the method to check your own estate.
Explore Our ResearchAI & Agent Security
Mitiga Labs Team
Read the Post →Identity & Access
Threat Hunting & Detection
Ucha Gobejishvili
Read the Post →AWS
Threat Hunting & Detection
Yafit Shayo
Read the Post →
INSTRUCTION FILES YOUR AGENT READS — AND NOBODY OPENS.
Rule files, hooks, and CLAUDE.md run automatically the moment an agent loads them. Mitiga Labs found instruction files quietly recording prompts, tokens, and environment variables, then shipping them out. No malware, often no audit log.
See How it WorksLet them come
A Research Report from Mitiga Labs


Mitiga Labs is where we take apart cloud, SaaS, identity, third-party services and AI attacks, then write what they find into the detections customers run. Those detections interrupt the next attack before it reaches the business. That's Zero-Impact Breach Prevention.
Explore the latest discoveries, deep dives, and technical how-tos from the Mitiga Labs team. Our research is dedicated to informing defenders and disrupting attackers.
A Mitiga Labs report on PromptLoggers and exfiltration hidden in AI instruction files — six real public-repo case studies, the techniques, and how to detect them before an agent runs.
Mitiga Research Team
Stolen ConnectWise app credentials gave attackers app-only access to Microsoft 365. A six-part method for hunting third-party app abuse in Entra ID.
Ucha Gobejishvili
Read the Post →In a security data lake, a detection can be correct and still never finish. How two months of PySpark tuning became 34 rules that cut runtime up to 60%.
Yafit Shayo
Read the Post →Identity & Access
Attack Research
Mitiga's research team uncovered a data risk to Okta users due to passwords that can be present in logs. This article outlines the risk and attack method.
Doron Karmi
Read the Post →Google Cloud
Threat Hunting & Detection
Hunt for complex data exfiltration activities inside Google Cloud environments using Mitiga's tactical security guide built for enterprise threat hunters.
Veronica Marinov
Read the Post →Google Cloud
Cloud Forensics & Logging
Learn how Mitiga addresses Google Cloud vulnerability management with enhanced forensic visibility for GCP Storage in this latest security advisory.
Veronica Marinov
Read the Post →Software Supply Chain & CI/CD
Threat Hunting & Detection
Learn how to investigate the CircleCI breach with Mitiga’s technical guide to assist organizational threat hunting efforts.
Doron Karmi
Read the Post →Software Supply Chain & CI/CD
In this blog, Mitiga Devops Engineer Stav Ochakovski addresses our organizational monorepo shift and why it triggered a CI adjustment, as well.
Stav Ochakovski
Read the Post →AWS
Attack Research
Read Mitiga research about a new post-exploitation attack method, a new way that enables adversaries to hijack public IP addresses for malicious purposes.
Or Aspir
Read the Post →AWS
Attack Research
A recent Mitiga Research Team investigation found the well-regarded Amazon Relational Database Service is leaking PII via exposed RDS Snapshots.
Ariel Szarf
Read the Post →Breach & Campaign Analysis
Cloud Forensics & Logging
On September the 16th, Uber announced they experienced a major breach in their organization in which malicious actor was able to log in and take over multiple services and internal tools used at Uber. What are some of the logs that IR teams should be focusing on in their investigation?
Or Aspir
Read the Post →Identity & Access
Breach & Campaign Analysis
Mitiga investigated an attempted Business Email Compromise (BEC) attack. While the alertness of the involved parties prevented the fraud, the attack indicated that the attacker had access to sensitive information that could only be obtained by compromising a user in the organization.
Mitiga's Research Team
Read the Post →
AI Detection and Response for modern infrastructure. Defend with AI, from AI, and your AI.
Learn more