blue-harness: an open-source binary anti-reverse-engineering hardening tool
Understand this story
SeifHeakal has released an open-source tool called blue-harness for performing exposure audits and hardening on one's own native binaries and mobile apps, and for verifying whether each protection actually triggers. The tool is open-sourced under the GNU AGPL-3.0 license and can be used free of charge, including by enterprises, provided the AGPL is complied with; evaluation and testing are free for everyone. The author ran a comparative test with a 4-bit quantized 7B local model: the unhardened version was cracked in 1.6 seconds, while the hardened version consumed 13 times the tokens and remained uncracked after about 195 seconds. However, the honeypot layer did not trigger, and that layer was not tested.
AI overview based on source reports · Updated 19 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIShow HN: Blue-harness – audit and harden binaries against AI-assisted reversing
blue-harness is an open-source, build-time anti-reverse-engineering hardening and audit harness for native binaries (Linux ELF, Windows PE, macOS/iOS Mach-O) that audits a build's exposure, applies nine countermeasures, and verifies each defense actually fires. In one measured run, a small 4-bit-quantised 7B model cracked the unprotected build in 1.6 s but failed against the hardened build after 195 s and 13× the tokens, though the honeytoken layer went untested. It is AGPL-3.0 licensed, blue-team only, and the author asks for testing against larger models via make slice-k-run MODEL=<your-model>.
Activity for this story
Not enough continuous observations to show a trend yet.