my-claude-skills

Triages and classifies program crashes by bug class and exploitability. Use when reviewing a crashing input, an AddressSanitizer report, a GDB/LLDB backtrace, or a core dump; when deciding whether a crash is a real memory-safety bug worth reporting or a benign input rejection; or when prioritizing a batch of fuzzer-found crashes. Reproduces a PoC under sanitizers/debugger and emits a severity-ranked verdict.

Analyzes network captures (pcap/pcapng) for digital forensics and reverse engineering. Use when examining packet captures, extracting network artifacts, investigating incidents, analyzing malware C2 traffic, or reconstructing network activity from capture files.

Analyzes binaries using angr's Python framework for static analysis, symbolic execution, and vulnerability detection. Use when decompiling executables, recovering control flow graphs, finding vulnerabilities via symbolic execution, or analyzing stripped binaries without source access.

Instruments running binaries using Frida's dynamic analysis toolkit. Use when tracing function calls at runtime, hooking APIs, inspecting process memory, enumerating loaded modules, or performing instruction-level tracing on live processes.