What problem does it solve?
Once a penetration test identifies a framework, component, or version, testers often jump straight to exploitation without checking public vulnerability intelligence, leading to blind attacks and missed known CVEs. This Skill enforces a mandatory, structured online reconnaissance routine before any exploit attempt.
Core Features & Use Cases
- Multi-source vulnerability search: Queries CVE databases (searchsploit, cve.circl.lu, GitHub advisories, CVEDetails), search engines (Google, Baidu, Bing, DuckDuckGo), and Chinese security communities (Seebug, FreeBuf, Qianxin, Anquanke) for known vulnerabilities and analysis.
- PoC and exploit discovery: Searches GitHub repositories and code via API for exploit code, plus asset engines (FOFA, Shodan, ZoomEye, Censys) and real-time channels (X, Reddit, Exploit-DB) for in-the-wild intelligence.
- Blocked-access fallback chain: When searches hit 403s, captchas, or empty results, it escalates through user-agent rotation, Jina reader, Google cache, Wayback Machine, GitHub API, alternate engines, and SOCKS5 proxies.
- Use Case: During an authorized engagement you fingerprint a target running Apache Struts 2.5.30. The Skill immediately searches CVE feeds, Chinese community write-ups, and GitHub for working PoCs, then recursively repeats the process for each dependency found in the component's manifest files.
Quick Start
Identify the target's framework and version, then ask the agent to run full vulnerability intelligence collection on that component before attempting exploitation.