Eresus research, advisory, and security news
We collect writing, advisories, and current-event analysis around AI security, the MCP ecosystem, application security, and real attack chaining here.
Latest Posts
What is Agentic Cybersecurity? Are AI Agents Replacing Pentesters?
Discover the differences between Agentic Security architecture and manual penetration testing, and explore the future of AI in offensive cybersecurity.
What Is AI Data Governance and Why Is It So Hard to Implement?
AI data governance guide for LLM and RAG systems: semantic DLP, tenant-aware retrieval, purpose-based access, provenance, deletion, and auditability.
How AI Coding Assistants Expose Secrets
AI coding assistant security guide for Cursor, Copilot, and Claude: secret leakage, workspace indexing, prompt injection, terminal tools, OAuth scopes, and enterprise controls.
Setting up an Active Directory Pentest Lab in Seconds with AI (Cursor)
For cyber security experts and internal Red Teams: How to set up a fully comprehensive, vulnerable Active Directory test environment with a single click...
Kubeflow Dashboard Security: Authentication, Notebook Isolation and Cloud Metadata Risk
How to assess Kubeflow dashboard exposure, OIDC, profiles, notebook permissions, network isolation, audit logs, and cloud metadata access without assuming every deployment is vulnerable.
PAIT-ARV-100: Archive Slip Vulnerabilities in Machine-Learning Model Packaging
Comprehensive guide on Zip Slip and Tar Slip path traversal in AI model archives, detailing how malicious archive entries escape extraction sandboxes and execute code on inference servers.
Beyond Jailbreaks: Contextual Red Teaming for Agentic AI
Why standalone jailbreak tests miss agent risk, and how to test indirect prompt injection, retrieval boundaries, tool permissions, and recovery in multi-step systems.