jongkwan.dev

Posts

15 posts
  1. RAG Datastore Membership Inference: The Retrieval Corpus as Attack Surface

  2. LLM Agent Memory Poisoning

  3. LLM Jailbreaks and Automated Red Teaming

  4. MCP Tool Poisoning

  5. LLM Privacy Defenses: DP, Encryption, TEEs, and Unlearning

  6. LLM Privacy Attacks: Training Data Extraction and Membership Inference

  7. Output Guardrails, Observability, and the Right to Erasure

  8. Egress Gateway Defense

  9. Defenses at the Agent, RAG, and MCP Layers

  10. Input-Side PII Anonymization and Reversible Mapping

  11. How Personal Data Leaks, and What Data Governance Has to Cover

  12. Exposing a Home Server Safely with Cloudflare Tunnel

  13. JWT Authentication and TOTP 2FA

  14. TLS and HTTPS: Handshakes and Certificates

  15. Kubernetes in Depth: Networking, Storage, and Deployment Strategies