2026-10-14
Day 14: Selective Memory
Lane: Binary reverse engineering. Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
binary-reverse-engineering-day-14-objective-c-persistence-descriptor - Prompt: "Objective-C never forgets: every class, every selector, every string is registered and waiting. Read the persistence descriptor and learn what this binary remembers."
What we were given
- One
evidence.zip(32,246,719 bytes, SHA-256cd745ce710566bf41873bc7708045490847cd91c5d4528fe75286b661ae6f10c). AES encrypted, passwordinfected. - The earlier files again:
d01tod13. - New today:
d14--variant-000.elf(17,304 bytes), a 64-bit Linux program written in Objective-C. The original note says it uses the GNUstep runtime; the file linkslibobjc.
What was asked
- 140 questions in Single Solve and Time Trial. Mastery had five deliveries, so 700.
- Question types:
- Hashes: SHA-256 of each member.
- d01 to d13: the same questions as Day 13.
- d14 (new): machine, architecture, format (lowercase token), the config fields, the CRC32, and
has_hcfg_section(true when the ELF section names include.hcfg).
How it was solved, start to finish
- Ran the driver
run_tm.py: start the attempt, download the zip, runsolve_tm.py, submit. - d01–d13 were answered as on Day 13.
- d14 config: the question said to scan for a literal
HCFGand accept the first valid record. Here it sits in its own.hcfgsection. Re-checked on the saved sample:
HCFG offset=4000 version=1 xor_key=0xa1 len=367 crc32=0x79f491da
section of match: .hcfg
JSON: {"campaign":"BR14-000","domain":"node-000.day14.example.invalid","ipv4":"198.51.100.20","url":"https://node-000.day14.example.invalid/api/BR14-000","mutex":"Global\\HuntressBR14_000", ...}
- d14 section check: the Day 10
has_hcfg_sectioncode reads the section names, so the answer wastrue. - Answers for d14: format
elf, architecturex86-64,has_hcfg_sectiontrue, CRC320x79f491da, campaignBR14-000. - Submitted all 140 answers (
answers.json). Result:CORRECT. - Time Trial: a fresh variant, same steps. Result:
CORRECT. - Mastery: five deliveries (700 answers), same steps. Result:
CORRECT.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | a1895cef4c7b9ba5d7eced20feb8432f |
| Time Trial | CORRECT | ee069c5160da6212762b442acdbf6c7d |
| Mastery | CORRECT | 14978c42a0caa6dbb550f8b8993d5269 |
Each mode folder has the raw start-response.json, answers.json, submit-response.json and final-status-response.json. The downloaded zips are under deliveries/.
Reproduce
Offline, from this folder (no Game Master needed):
mkdir -p /tmp/d14/deliveries && cp single-solve/evidence.zip /tmp/d14/deliveries/delivery-1.zip
python3 -I ~/.config/huntress/solvers/solve_tm.py single-solve/start-response.json /tmp/d14/deliveries /tmp/d14/answers.json
Checked on 2026-11-01: this gives the same 140 answers as answers.json. It only parses the samples and never runs them.
Live driver:
python3 -I ~/.config/huntress/solvers/run_tm.py binary-reverse-engineering-day-14-objective-c-persistence-descriptor single_solve /tmp/d14
This needs a Game Master token in ~/.config/huntress/gm-token.
Day 14: The Shortcut Home
Lane: Metadata forensics. Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
metadata-forensics-day-14-shortcut-home - Prompt: "A shortcut collection is a map of intent—if you read where each link points instead of where it sits. Reconstruct the targets and identify the impossible destination."
What we were given
- Single Solve:
(evidence.zipmetadata-day-14-evidence.zip, 475 bytes, SHA-256cd5858414bfacebf40f6668b05456ff36d7626847737b5b7f0cd2d5e9745edec) withshortcut-001.lnk. - Time Trial:
evidence.zip(604,993 bytes, SHA-25610c59cb8f45d5aa23ba931688fe73b5f6765bddc94638ec4a70c8684cebaf874). One file of each type so far, now includingshortcut-001.lnk. - Mastery:
metadata-forensics-day-14-shortcut-home-mastery.zip(3,025,411 bytes, SHA-25678b621838abb63d52c35159d57901f845cfdce0a11f59abdfafcdc4a849a1a9a). Five inner zips,evidence-01.ziptoevidence-05.zip. - New today: Windows shortcut (
.lnk) files.
What was asked
- Single Solve: 1 question:
lnk.target_path_windowsforshortcut-001.lnk. - Time Trial: 19 questions. Mastery: 95 (5 sets × 19).
- Question types: everything from Day 13 (file identity, image/PDF/ZIP headers, PE fields, .NET MVID, MSI
ProductCode, SHA-256 ordering), plus the new LNK target path.
How it was solved, start to finish
The method: a .lnk file stores its target in the LinkInfo block as a local base path. Read it from the bytes.
- Wrote a small LNK parser,
lnk_target.py: - Check the header: the file starts with
4C 00 00 00. - Read the
LinkFlagsat0x14. - If the flag
HasLinkTargetIDList(0x1) is set, skip the ID list (2-byte size + data) that starts at0x4C. - If
HasLinkInfo(0x2) is set, read theLinkInfoblock. When itsVolumeIDAndLocalBasePathflag is set, read theLocalBasePathstring at the stored offset. - Add the
CommonPathSuffixstring. If theLinkInfoheader is0x24bytes or more, prefer the Unicode base path.
flags = struct.unpack_from('<I', b, 0x14)[0]
pos = 0x4c
if flags & 0x1: pos += 2 + struct.unpack_from('<H', b, pos)[0] # skip IDList
if flags & 0x2:
li = pos
lbp_off = struct.unpack_from('<I', b, li + 0x10)[0] # LocalBasePath
cps_off = struct.unpack_from('<I', b, li + 0x18)[0] # CommonPathSuffix
- Tested it on the Single Solve file.
shortcut-001.lnk→C:\Windows\System32\cmd.exe.CORRECT. - Wired it into
run_meta.py. Answer IDs withlnk-targetcalltarget_path(). All other fields used the existing handlers. - Time Trial and Mastery.
run_meta.pyanswered all 19 and all 95. The LNK target wasC:\Windows\System32\cmd.exein every set. Example synthesis answers: Time Trial["document-001.pdf","shortcut-001.lnk","evidence-note.txt"], mastery set 1["shortcut-001.lnk","evidence-note.txt","document-001.pdf"]. BothCORRECT.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | dd433920a61ec2c8b76ce90c1a09744e |
| Time Trial | CORRECT | 4766855aec3314855c0a2174a2a3591e |
| Mastery | CORRECT | d1567c6181898e2102a2822f6c9a7718 |
Each mode folder has answers.json, start-response.json, submit-response.json and final-status-response.json. The Single Solve folder also has the zip. No screenshots were captured.
Reproduce
python3 -I ~/.config/huntress/solvers/lnk_target.py shortcut-001.lnk
python3 -I ~/.config/huntress/solvers/run_meta.py metadata-forensics-day-14-shortcut-home single_solve /tmp/d14
The first line only needs the extracted .lnk. The second needs a Game Master token in ~/.config/huntress/gm-token; for time_trial and mastery it calls op=retry first, which spends one of the unique attempts. Nothing here opens the shortcut in Windows.
Day 14: Predictable Crypto
Lane: Mobile (Android). Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
mobile-android-day-14-predictable-crypto - Prompt: "Exploit seeded nonces or weak derivation without bypassing earlier storage gates."
What we were given
- One Android APK per case. Inside:
AndroidManifest.xml,classes.dex,assets/huntress_puzzle.json, andMETA-INF/signing files. - Single Solve:
predictable_crypto.apk, caseMOB-14-B14DCCA3D9(12,704 bytes, SHA-256b6d3fff1c44485eab54e3b8fb89ad5d0a30e509bdc2b60dd35fbcca0b7400fee). Its puzzle is saved assingle-solve/huntress_puzzle.json. - Time Trial: one APK (case
MOB-14-BC19FA2A71). Mastery: five APKs. These APKs were not saved in the repo.
What was asked
- One question per case: return the exact backend answer code for that case.
- Answer format:
ANS-plus exactly 16 uppercase hex characters. - Answer IDs:
answerfor Single Solve and Time Trial.set1.answertoset5.answerfor Mastery. - Mastery mixed cases from Days 2, 4, 7, 10 and 14.
How it was solved, start to finish
The app was never installed or run.
- Started the Single Solve and downloaded the APK.
- Read the puzzle asset. The codec is
caesarwith key 9 (Day 7 used key 10).
{"schema":"huntress.mobile.puzzle.v1","day":14,"codec":"caesar","data":"JW\\6L><BL=A;OL>NN:=?","key":9}
- Decoded it. Shift every character back by
key. For exampleJ(74) becomesA(65).
import json
p = json.load(open('single-solve/huntress_puzzle.json'))
print(''.join(chr((ord(c) - p['key']) % 256) for c in p['data']))
Output: ANS-C539C482FC5EE146.
- Ran the lane driver for all three modes:
python3 -I ~/.config/huntress/solvers/run_mode.py mobile-android-day-14-predictable-crypto <mode> mobile_puzzle <workdir>
Single Solve and Time Trial came back CORRECT.
- Fixed a Mastery crash. Mastery crashed inside the Base64 step. Some sets carried Base64 that was not clean (padding and other malformed input). The Base64 decode in
mobile_puzzle()was hardened, then Mastery was run again. The hardened lines inrun_mode.py:
dd=''.join(pj['data'].split()).replace('-','+').replace('_','/') # strip ws, url-safe
if len(dd)%4==1: dd=dd[:-1] # drop stray trailing char
raw=base64.b64decode(dd + '='*(-len(dd)%4), validate=False)
It strips whitespace, accepts URL-safe characters, drops a stray last character, and adds missing = padding.
- Results: Single Solve (
ANS-C539C482FC5EE146), Time Trial (ANS-BDBD1DDEEF08BECB) and all five Mastery answers came backCORRECT.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | 3adc5ad417817e22db9226cdf0f10cbd |
| Time Trial | CORRECT | 5dc49487045a8e3ec3a1b9ff0d2268af |
| Mastery | CORRECT | 227c27f165ce1fba33b1cfb409a92b09 |
Each mode folder has answers.json, start-response.json, submit-response.json and final-status-response.json. No screenshots were taken.
Reproduce
Offline: run the Python snippet above from this folder.
A fresh run with the driver:
python3 -I ~/.config/huntress/solvers/run_mode.py mobile-android-day-14-predictable-crypto single_solve mobile_puzzle /tmp/d14
This needs a Game Master token in ~/.config/huntress/gm-token. It only unzips the APK and reads the asset. It never installs or runs the app.
Day 14: Trust the Process
Lane: Script analysis. Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
script-analysis-day-14-msbuild-trusted-project-abuse - Prompt: "An MSBuild project looks like routine build plumbing until you follow its properties into its targets. Recover the property and target flow, compare it with earlier execution chains, and find where a trusted toolchain is doing untrusted work."
What we were given
- One
evidence.zip(103,827 bytes, SHA-25698b763891adf3fabc82a97dbec0cb2db4e24c44c68f15df09607997af0e07c6d). Password:infected. - The same families as Day 13. New today:
- 4 MSBuild project polyglots:
polyglot-msbuild_xml-benign-utility.proj,-downloader,-launcher,-staged-chain. - 2 more advanced Bash files:
brace-expansion.sh,getopts-state.sh. - Every host and polyglot file carries the
HUNTRESS_ANALYSIS_V1:<ENC>-<data>marker fromDay 8.
What was asked
- 971 questions in Single Solve. By family: PowerShell 15, JScript 72, Batch 72, VBScript 72, Python 126, Bash 140, Node.js 56, host 66, polyglot 352.
- The question types are the same as before. Host and polyglot files get 11 questions each (role, ATT&CK, adapter, campaign, URL or path, host profile, obfuscation, stage profile, next language, stage count, steps).
- Time Trial: 971 questions. Mastery: 4,855 questions in 5 sets.
How it was solved, start to finish
- Started the untimed Single Solve and downloaded
evidence.zip. Extracted it with7z x -pinfected. - Ran the existing solver (
solve_expansion.py, driven byrun_expansion.py). Each family has its own decoder:
| Family | Files | How the config is recovered | First seen | |
|---|---|---|---|---|
| PowerShell | *-sample.ps1 | Base64 JSON inside FromBase64String('...') | Day 1 | |
| JScript | *.js | jsdec.js loads the fixture in a Node vm sandbox (pure-computation globals only) and calls its decode functions | Day 3 | |
| Batch | *.bat, *.cmd | joins the :HUNTRESS_INSPECT fragments and drops ^ carets | Day 3 | |
| VBScript | *.vbs | Material_ + Codec_, transform search until the text starts `v=1\ | steps=` | Day 4 |
| Python | *.py | MATERIAL_ read with ast.literal_eval, same transform search | Day 5 | |
| Bash | *.sh | quoted Material_ + Codec_; "advanced" files start `v=1\ | feature=` | Day 6, Day 7 |
| Node.js | node-*.mjs | Material_ + Codec_ with \x, \u and hex-array formats | Day 7 | |
| host, polyglot | host-*, polyglot-* | HUNTRESS_ANALYSIS_V1:<ENC>-<data> marker (B64, HEX, HB64, FRAG) | Day 8 |
Role comes from the ordered steps, and ATT&CK from the role and the language (rules in Day 3). Host and polyglot configs already contain their role and ATT&CK ID.
- No new code was needed. The MSBuild
.projfiles carry the usual marker.
- Ran
run_expansion.pyfor all three modes. All three came backCORRECT.
- Real answers from
single-solve/answers.json:
{
"polyglot-msbuild_xml-staged-chain-role": "staged_chain",
"polyglot-msbuild_xml-staged-chain-attack": "T1059",
"polyglot-msbuild_xml-staged-chain-adapter": "operation_graph",
"polyglot-msbuild_xml-staged-chain-campaign": "MSBUILD_XML-984e0b62",
"polyglot-msbuild_xml-staged-chain-url": "https://msbuild_xml-984e0b6296.example.invalid/stage",
"polyglot-msbuild_xml-staged-chain-stage-count": "2",
"polyglot-msbuild_xml-staged-chain-steps": "acquire_remote>dispatch_process",
"polyglot-msbuild_xml-staged-chain-host-profile": "standalone"
}
- Time Trial and Mastery used the same solver on fresh variants.
- Time Trial: 971 questions, no deadline (
run_for_time: false). - Mastery: 4,855 questions in 5 sets, no deadline (
run_for_time: false). The download (script-analysis-day-14-msbuild-trusted-project-abuse-mastery.zip) holds one inner zip per set. The driver solves each set and adds thesetN.prefix. - Both came back
CORRECT.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | d00c06c24a63ee4f1dafcdabbcc583e2 |
| Time Trial | CORRECT | 44e6cc3d28acf695d5af568bb6bc2929 |
| Mastery | CORRECT | 4cca9097328a024d16ef5ddd690904f9 |
Each mode folder has answers.json and the Game Master JSON (start-, submit-, final-status-response.json). The evidence archives were not kept. No screenshots were taken.
Reproduce
The evidence archive was not kept, so the saved answers cannot be rebuilt offline. For a fresh attempt of this format:
mkdir -p /tmp/d14/ex
7z x -pinfected -o/tmp/d14/ex evidence.zip
python3 -I tools/solve_expansion.py /tmp/d14/ex start-response.json /tmp/d14/answers.json
Then submit answers.json to the Game Master (needs a token in ~/.config/huntress/gm-token). The tools/ copy decodes JScript with the pure-Python jsdec_py_emu.py, not the Node sandbox.
On the day, run_expansion.py drove the live modes. Since Day 23 that driver sends any question set with polyglot- IDs to solve_day23.py, which does not handle this format's host- and resource- IDs. So use solve_expansion.py directly, as above.
Day 14 — Seven Steps (Caesar Cipher Variant)
Lane: Warmups · Challenge ID: warmups-day-14
> Caesar took a week off.
What we were given
A text blob encoded with a shifted cipher. Each blob uses a different Caesar shift value.
How it was solved
"Caesar took a week off" — a week is 7 days, so this is a Caesar cipher with shift 7 (ROT7 instead of the usual ROT13).
- Downloaded the blob from the Game Master.
- The blob contains fragments that are additionally transformed with per-fragment ciphers (reverse, atbash, or further Caesar shifts).
- Auto-detected the shift value by trying all 26 rotations and scoring against English letter frequency / known slug vocabulary.
- Decoded each fragment, reassembled the plaintext, and extracted the answer slug.
For mastery, multiple blobs were decoded independently. Slug words were disambiguated against the observed CTF vocabulary when the shift produced near-matches.
Automated with warmup_decode.py.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | 13d100f8ceb18f836e51952aec06edee |
| Time Trial | CORRECT | 7b42fabbb3290101ebc3061c34182ad4 |
| Mastery | CORRECT | f43bb754e48931c07dfcb4dcfea61fa6 |