The challenge
A "control-flow reconstruction" day: every earlier day's binary fields plus a new Day 18 binary. Written from our solver's report and the start reply's question list; the binaries weren't reopened.
Safety
Binaries are hostile: static analysis only, on our Kali VM. Nothing executed, no excerpts quoted.
What was in the files
216 questions on the Single Solve. The 12 new d18_* fields were all familiar families: the artifact's SHA-256, format, architecture and machine, the HCFG config (campaign, domain, mutex, URL, CRC32), the HCFG offset and version, and the XOR key. Mastery came as 4 separate downloads.
The modes
- Single Solve: core batch of 183 accepted, then one merged plug-in batch of 33 → CORRECT at 09:01:02. No ranked candidates for format or architecture were needed.
Single Solve CORRECT 09:01:02
Time Trial CORRECT 09:01:13 flag 400c9d34…7103 solve 733 ms, submit +3.99 s
Mastery CORRECT 09:01:23 flag 5e61d8a5…3f77 solve 354 ms, submit +3.43 s (4 downloads streamed)
Time to solve
All three modes were solved by 09:01:23 ET, about a minute and a half after release, each on the first submit and with no hand work.
Takeaways
- The fallback plan wasn't needed today: we had a plan for the Homebase agent (GPT-6 Luna) to work new Binary fields, but the solver's chain covered every new field by itself.