Research Directions
Focused on real-world offense and defense — Web security, cloud security and agentic security.
01 / PENETRATION TEST
Penetration Testing
Studying the weak points of target systems from an attacker's perspective — finding and proving issues before real risks materialize.
- Web application and API penetration research
- Intranet lateral movement and AD techniques
- Mini-program, mobile and cloud-specific offense
- Vulnerability hunting and SRC engagements
02 / CLOUD SECURITY
Cloud Security Research
Research on attack surfaces and exploitation across major cloud providers — distilled into open-source tools like lc, cloudsword and CF.
- Multi-cloud attack surface inventory and mapping
- Full post-credential cloud exploitation workflows
- Cloud tenant risk discovery and testing
- Open-sourcing cloud security tools
03 / AGENTIC SECURITY
Agentic Security Research
Exploring the boundaries of AI agents in offense and defense — a full pipeline from structured knowledge to autonomous pentest agents.
- Agent-executable structured pentest knowledge base (AboutSecurity)
- Knowledge-as-API via MCP (context1337)
- Autonomous pentest agent with multi-round decisions (tchkiller)
- Agent attack & defense evaluation arena (HunXiang)
04 / CODE AUDIT
Code Audit
Digging into source code to uncover security risks — eliminating vulnerabilities before they ship.
- Audits for Java / PHP / Python / Go and more
- Deep discovery of business-logic flaws
- Third-party component and supply-chain risk review
- Root-cause research and write-ups