blog category
AI & Security
AI & Security
Accelerating EDR Evasion with LLM-Driven Analysis
Over the years I have enjoyed disassembling and debugging endpoint detection and response (EDR) and antivirus...
By: Adam Chester
Jun 29, 2026 • 17 min read
Read Post
AI & Security
Jailbreaker: LLM Jailbreak Testing You Can Actually Repeat
You may have read about our new GhostWorks initiative here at SpecterOps. As part of this...
By: Neeraj Gupta
Jun 29, 2026 • 12 min read
Read Post
AI & Security
Time Travel Debugging with Codex
Spoiler warning: this post discusses the FLARE-ON 12 `FlareAuthenticator` challenge and includes the recovered flag. Introduction...
By: Kai Huang
Jun 26, 2026 • 9 min read
Read Post
AI & Security
Disposable Tooling: Building LLM-Generated Mythic Agents from Prompt to Deployment
An area I’ve been very interested in exploring over the past several months is the generation...
By: Adam Chester
Jun 24, 2026 • 25 min read
Read Post
AI & Security
SpecterOps and OpenAI: Helping to Build a New Security Frontier with Daybreak
Today, OpenAI announced that it is expanding access to its frontier AI cybersecurity capabilities for a...
By: Jared Atkinson
Jun 22, 2026 • 6 min read
Read Post
AI & Security
BloodHound MCP, One Year Later: What I Learned About MCPs, Models, and Context
The first version of BloodHound MCP proved that an LLM could converse with BloodHound. The current...
By: Matthew Nickerson
Jun 18, 2026 • 15 min read
Read Post
AI & Security
Building an Indirect Prompt Injection Workflow
TL;DR: This post covers how I used OpenAI’s Codex to automate the generation, testing, and refinement...
By: Antero Guy
Jun 11, 2026 • 26 min read
Read Post
AI & Security
Oops, I Weaponized the Database: Abusing AI Features in SQL Server 2025
TL;DR: Microsoft SQL Server 2025 AI features provide a practical channel for data exfiltration and C2...
By: Justin Kalnasy
Jun 10, 2026 • 21 min read
Read Post
AI & Security
Prompt Engineering for Security Agents: A Measurable Approach with GEPA
You may have read about our new GhostWorks initiative here at SpecterOps. As part of this...