What belongs in AI-accelerated security research?
SuperDevSecurity is a public home for research that uses AI systems as part of the security research process, not as a replacement for technical judgment. The work here should be testable, useful, and honest about where automation helps or fails.
Scope
The site will prioritize work that can be inspected or reused:
- Methodology for AI-assisted code review, vulnerability discovery, exploit reasoning, and triage.
- Reproducible reports with clear assumptions, inputs, prompts, tooling, and failure modes.
- Public tools that make security research faster without hiding the evidence chain.
Standard
Published work should make a concrete claim and carry enough detail for another researcher to challenge it. That means preserving artifacts, naming model/tool versions when relevant, and separating confirmed findings from promising leads.
Direction
The initial focus is practical: agentic research workflows, vulnerability research acceleration, secure use of AI in engineering organizations, and public tooling that helps teams evaluate these systems with less guesswork.