2026-10-18
Day 18: Flow State
Lane: Binary reverse engineering. Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
binary-reverse-engineering-day-18-control-flow-reconstruction - Prompt: "Follow the calls and the program tells you its story. Reconstruct the call graph, read the API intent at each edge, and prove no injection hides in the flow."
What we were given
- One
evidence.zip(32,835,227 bytes, SHA-2562e8dcfbe27cb788688817c9a43f5cd41d072f22e38dbcf0756fcc76e0c6a9dad). AES encrypted, passwordinfected. It was not saved in this folder. - The earlier files again:
d01tod17. - New today:
d18--variant-000.elf, a small 64-bit Linux program (SHA-25600882cd984852d5cc186662a6cb7859c1cf465f7af60349bc5ffe89e3543a98e). - Its names include
event_success,event_decoy,event_failure,event_sandboxandmodeled_process_access_no_injection. - The same file (same SHA-256) is inside the Day 19
evidence.zip, so it can still be checked. - The bundles are now about 33 MB each.
What was asked
- 216 questions in Single Solve and Time Trial. Mastery had four deliveries, so 864, under a timer.
- Question types:
- Hashes: SHA-256 of each member.
- d01 to d17: the same questions as Day 17.
- d18 (new): the standard fields only: machine, architecture, format, XOR key,
HCFGversion,hcfg_offset, the config fields and the CRC32. No call-graph questions were asked.
How it was solved, start to finish
- Ran the driver
run_tm.py: start the attempt, download the zip, runsolve_tm.py, submit. - Made the downloads robust (the fix for this day). With ~33 MB bundles, a download could arrive cut short.
run_tm.pynow:
- compares the bytes received with the
Content-Lengthheader and retries a short read; - checks that the zip's central directory opens (a full download), and downloads again if not;
- tries up to 6 times.
- d01–d17 were answered as on Day 17.
- d18 config: a literal record, found by the normal scan. Re-checked on the copy from the Day 19 bundle with
hcfg.py:
=== HCFG valid records in d18--variant-000.elf ===
offset=0x2140 version=1 xor_key=0x31 plen=367 crc32=0xd55cf099
JSON: {"campaign":"BR18-000","domain":"node-000.day18.example.invalid","ipv4":"198.51.100.90","url":"https://node-000.day18.example.invalid/api/BR18-000","mutex":"Global\\HuntressBR18_000", ...}
0x2140 is 8512. Answers: format elf, XOR key 0x31, version 1, hcfg_offset 8512, campaign BR18-000.
- Submitted all 216 answers (
answers.json). Result:CORRECT. - Time Trial: a fresh variant, same steps. Result:
CORRECT. - Mastery: four deliveries (864 answers), same steps. Result:
CORRECT.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | bae5ac55252bd7a7c3605dc49d07453a |
| Time Trial | CORRECT | 3d99da461084ca1878d7d089dc31bbec |
| Mastery | CORRECT | 082390eb12c9ca5eb65914306f4e7855 |
Each mode folder has the raw start-response.json, answers.json, submit-response.json and final-status-response.json. The evidence zips were not saved for this day.
Reproduce
The Day 18 bundle was not saved, so a full offline re-run is not possible. The d18 config can be checked from the Day 19 bundle (from this folder):
mkdir -p /tmp/d18 && 7z x -y -pinfected -o/tmp/d18 ../day-19-binary-jack-in-the-box/single-solve/evidence.zip 'd18--*'
sha256sum /tmp/d18/d18--variant-000.elf # 00882cd9…, same as d18_artifact_sha256_1
python3 -I ~/.config/huntress/solvers/hcfg.py /tmp/d18/d18--variant-000.elf
Live driver:
python3 -I ~/.config/huntress/solvers/run_tm.py binary-reverse-engineering-day-18-control-flow-reconstruction single_solve /tmp/d18
This needs a Game Master token in ~/.config/huntress/gm-token. Both only parse the files and never run them.
Day 18: Frame of Reference
Lane: Metadata forensics. Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
metadata-forensics-day-18-frame-of-reference - Prompt: "Clips, previews, and screen recordings have been mixed together. Use stream geometry to identify originals, crops, and the recording that could not have come from the claimed device."
What we were given
- Single Solve:
(evidence.zipmetadata-day-18-evidence.zip, 18,665 bytes) withvideo-001.mp4. - Time Trial:
evidence.zip(626,852 bytes, SHA-25613130400493ee751e42950d040a7803056845bb360e7b4bbd2a39d0fb7040ec6). One file of each type so far, now includingvideo-001.mp4. - Mastery:
metadata-forensics-day-18-frame-of-reference-mastery.zip(3,134,684 bytes, SHA-256e2d098de75bb3e09521ad5199185937dc241a2027a23f0e9b6f9b5a0b8acf05b). Five inner zips,evidence-01.ziptoevidence-05.zip. - New today: video files (
.mp4).
What was asked
- Single Solve: 1 question:
media.primary_video_dimensionsforvideo-001.mp4. Answer form:WIDTHxHEIGHT. - Time Trial: 23 questions. Mastery: 115 (5 sets × 23).
- Question types: everything from Day 17, plus the size of the first video stream.
How it was solved, start to finish
The method: take the width and height of the first video stream from the container, using ffprobe.
- Added a second
ffprobehandler torun_meta.py. It asks for streamv:0and printswidthandheightjoined byx:
r = subprocess.run(['ffprobe', '-v', 'quiet', '-select_streams', 'v:0',
'-show_entries', 'stream=width,height', '-of', 'csv=p=0:s=x', tf.name],
capture_output=True, text=True, timeout=30)
return r.stdout.strip() or None
- Single Solve.
video-001.mp4→320x180.CORRECT. - Time Trial and Mastery. Helper agents ran
run_meta.py. It answered all 23 and all 115. The video size was320x180in every set. Example synthesis answer (Time Trial):["video-001.mp4","evidence-note.txt","installer-001.msi","panel-001.png"]. BothCORRECT.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | 3364ee636ea58aaeb4893495266645fd |
| Time Trial | CORRECT | 757f8651706f9fb064d8fba77c6b5b47 |
| Mastery | CORRECT | 68d2e2aa477fd2a6bf55cfa4c26abc3f |
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; its saved submit says already_solved because it is a re-run, and the final status shows CORRECT. No screenshots were captured.
Reproduce
ffprobe -v quiet -select_streams v:0 -show_entries stream=width,height -of csv=p=0:s=x video-001.mp4
python3 -I ~/.config/huntress/solvers/run_meta.py metadata-forensics-day-18-frame-of-reference single_solve /tmp/d18
The second line 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. ffprobe only reads the container headers.
Day 18: Encrypted Strings
Lane: Mobile (Android). Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
mobile-android-day-18-encrypted-strings - Prompt: "Recover only the strings that reach security-sensitive Android and backend APIs."
What we were given
- One Android APK per case. Inside:
AndroidManifest.xml,classes.dex,assets/huntress_puzzle.json, andMETA-INF/signing files. - Single Solve:
encrypted_strings.apk, caseMOB-18-E2C0B7B304(12,704 bytes, SHA-2563df51a8bdf3790cd5b9d93f57fc0869fcb3ba07a66d85b27e504b8d56443367e). Its puzzle is saved assingle-solve/huntress_puzzle.json. - Time Trial: one APK (case
MOB-18-52DD746797). 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 3, 7, 12, 13 and 18.
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
xor, this time with key 20.
{"schema":"huntress.mobile.puzzle.v1","day":18,"codec":"xor","data":"VVpHOSZVJyAjIldRJSwjIyJXJ1c=","key":20}
- Decoded it: Base64-decode
data, then XOR every byte withkey.
import json, base64
p = json.load(open('single-solve/huntress_puzzle.json'))
print(bytes(b ^ p['key'] for b in base64.b64decode(p['data'])).decode())
Output: ANS-2A3476CE18776C3C.
- Ran the lane driver for all three modes. It already handled
xor:
python3 -I ~/.config/huntress/solvers/run_mode.py mobile-android-day-18-encrypted-strings <mode> mobile_puzzle <workdir>
- Results: Single Solve (
ANS-2A3476CE18776C3C), Time Trial (ANS-8889E945197A9A9E) and all five Mastery answers came backCORRECT.
Results
| Mode | Result | Flag |
|---|---|---|
| Single Solve | CORRECT | 9d051ec3a0a1eff6f74952577dd6e766 |
| Time Trial | CORRECT | dd78bef57e1bdc88d0433034788fa60b |
| Mastery | CORRECT | f2902f8beb9dd5774c335d23b666c86c |
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-18-encrypted-strings single_solve mobile_puzzle /tmp/d18
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 18: Orphan Process
Lane: Script analysis. Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
script-analysis-day-18-process-story - Prompt: "Source code describes which process should spawn which, and native telemetry keeps its own record of events. Reconstruct the static process tree, and catch the child claiming a parent it never had."
What we were given
- One
evidence.zip(119,408 bytes, SHA-256b75763de750f6378553db861950aacb11b1f7d2832c6a8183b351e35f3ade48d). Password:infected. - The same families as Day 17. New today:
- 1 new host file:
host-google-docs-sidebar-google-apps-script.gs(a Google Docs sidebar). - 4 Google Apps Script polyglots:
polyglot-google_apps_script-benign-utility.gs,-downloader,-launcher,-staged-chain. - A new marker form, SPLIT. All five new
.gsfiles use it instead of the one-line marker. - Every host and polyglot file carries the
HUNTRESS_ANALYSIS_V1:<ENC>-<data>marker fromDay 8.
What was asked
- 1,194 questions in Single Solve. By family: PowerShell 15, JScript 72, Batch 72, VBScript 72, Python 126, Bash 154, Node.js 56, host 99, polyglot 528.
- 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: 1,194 questions. Mastery: 5,970 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.
- The Single Solve crashed. Some host and polyglot files had no
HUNTRESS_ANALYSIS_V1:<ENC>-marker. Instead they had aHUNTRESS_ANALYSIS_V1:SPLITcomment and an array of string chunks. Real lines frompolyglot-google_apps_script-downloader.gs:
const HUNTRESS_ANALYSIS_V1_SPLIT_22eaee83 = [
"FRAG-763d337c6c616e6775616765.3d676f6f676c655f61707073.5f7363726",
"970747c726f6c65.3d646f776e6c6f616465727c.766572646963743d7375737",
...
- Fixed
hostpgdec.py. Join the chunks in order. The result is the normal<ENC>-<data>payload (hereFRAG-...), which then decodes as before:
am = re.search(r'HUNTRESS_ANALYSIS_V1_SPLIT_\w+\s*=\s*\[(.*?)\]', text, re.S)
joined = ''.join(re.findall(r'"([^"]*)"', am.group(1))) # "FRAG-763d337c..."
enc, data = re.match(r'([A-Z0-9]+)-(.+)$', joined, re.S).groups()
Decoded, the Google Docs host gives:
v=3|language=google_apps_script|role=staged_chain|verdict=suspicious|attack=T1059|adapter=operation_graph|host_profile=google_docs_sidebar|obfuscation_profile=fragmented_hex|stage_profile=cross_language_handoff|next_language=jscript|stage_count=3|steps=profile_host,decode_configuration,dispatch_language|...
- Submitted. The Single Solve was
CORRECT.
- The timed modes first got HTTP 409. Many lanes were running at once and an attempt was still active. A re-run went through.
- Real answers from
single-solve/answers.json:
{
"host-google-docs-sidebar-google-apps-script-role": "staged_chain",
"host-google-docs-sidebar-google-apps-script-attack": "T1059",
"host-google-docs-sidebar-google-apps-script-adapter": "operation_graph",
"host-google-docs-sidebar-google-apps-script-campaign": "GOOGLE_APPS_SCRIPT-29193ae3",
"host-google-docs-sidebar-google-apps-script-url": "https://google_apps_script-29193ae3e0.example.invalid/stage",
"host-google-docs-sidebar-google-apps-script-stage-count": "3",
"host-google-docs-sidebar-google-apps-script-steps": "profile_host>decode_configuration>dispatch_language",
"host-google-docs-sidebar-google-apps-script-host-profile": "google_docs_sidebar"
}
- Time Trial and Mastery used the same solver on fresh variants.
- Time Trial: 1,194 questions, no deadline (
run_for_time: false). - Mastery: 5,970 questions in 5 sets, no deadline (
run_for_time: false). The download (script-analysis-day-18-process-story-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 | 5aecf682b19d394d0160295a0886e3d8 |
| Time Trial | CORRECT | 6212f301fb57dfde703bea47d3e8adc5 |
| Mastery | CORRECT | 050a663b95a0252f722562745b602520 |
Each mode folder has answers.json and the Game Master JSON. The Single Solve folder also keeps evidence.zip. No screenshots were taken.
Reproduce
Offline, from the repo root:
mkdir -p /tmp/d18/ex
7z x -pinfected -o/tmp/d18/ex repos/day-18-script-analysis-orphan-process/single-solve/evidence.zip
python3 -I tools/solve_expansion.py /tmp/d18/ex \
repos/day-18-script-analysis-orphan-process/single-solve/start-response.json /tmp/d18/answers.json
This rebuilds all 1,194 saved answers (re-checked on 2026-11-01). It only parses the files. 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 18: Off the Rails
Lane: Warmups. Modes: Single Solve, Time Trial, Mastery. Result: all CORRECT.
The challenge
- ID:
warmups-day-18 - Prompt: "Two rails, one message. Read the fence."
- New encoding: rail fence cipher (2 rails)
What we were given
- One blob file per mode (Single Solve and Time Trial), or a zip of multiple blobs (Mastery).
- Each blob has numbered lines (
NN<TAB>fragment). Each line is independently encoded. - The decoded text ends with a slug like
word-word-word-NNto submit.
How it was solved
- Decode the blob. Characters are interleaved across 2 rails. First half of the string is the top rail (even positions), second half is the bottom rail (odd positions). Re-interleave to decode.
- Mastery blobs use multiple encodings from Days 1-18. The solver (
warmup_decode.py) tries every known decoder on each line and picks the one producing printable text. - Concatenate all decoded lines. The text contains a message ending in the answer slug.
- Submit the slug to the Game Master.
Results
- Single Solve:
CORRECT. slug:dapper-summit-brook-29. Flag:0a964fed6ed294c7bc81d6944b1c4d9a. - Time Trial:
CORRECT. slug:handy-pillow-juniper-48. Flag:d37d44147d1bc5966f4eab8b2df2194a. - Mastery:
CORRECT. 88 slugs. Flag:28514090552d16749534af35f1adee95.