What problem does it solve?
When a native binary crashes on oversized input during an authorized assessment, it is hard to tell whether the crash is a mere denial of service or a controllable return-address overwrite. This Skill provides a repeatable methodology to triage crashes, compute exact overwrite offsets, and build working exploits based on the binary's mitigations.
Core Features & Use Cases
- Crash Triage and Offset Discovery: Use De Bruijn cyclic patterns with gdb/pwndbg/GEF and pwntools to confirm instruction-pointer control and compute the exact offset to the saved return address.
- Mitigation-Aware Exploitation: Select ret2win, stack shellcode, ret2libc, or ROP chains based on checksec results (canary, NX, PIE, RELRO), including stack-alignment fixes and bad-byte enumeration.
- Advanced Scenarios: Covers Windows SEH overwrites, partial pointer overwrites, and byte-by-byte canary brute forcing against forked servers.
- Use Case: During a CTF or authorized pentest, a 64-bit non-PIE binary crashes on long input; use this methodology to confirm RIP control at offset 72 and build a ret2win payload that calls an unused win() function.
Quick Start
Ask the AI to analyze a crashed vulnerable binary with checksec and a cyclic pattern, determine the return-address offset, and build a pwntools exploit payload appropriate to its mitigations.