posted too much: a profile update that binds role fields

huntress ctf 2026, day 23 · web exploitation · single, time trial, and mastery

tools used: live web target capture and saved web inspection.

the visible form appeared to edit a profile name, but the json endpoint accepted additional role fields. the target's audit route exposed whether those fields changed the role seen by the application.

challenge shape

each assigned target supplied its own clue routes and state. mastery could mix this profile application with other web families, so the base title was not enough to classify every target in the bundle.

artifact anatomy

/api/clues identified NestJS and paired a feature route with a secondary audit route. the audit returned the exact json error {"error":"administrator_required"} before the update. that response was the guard for the profile flow.

extraction method

read the audit first. only after that exact error, send a synthetic profile update using the documented method and body shape. the bounded request used name and role fields accepted by this target, then read the audit again for the proof.

clues = get_json("/api/clues")
audit = get_json(clues["secondary_route"])
if audit == {"error": "administrator_required"}:
    post_json(clues["feature_route"], synthetic_profile_with_role_fields())
    proof = read_proof(get(clues["secondary_route"]))

validation and answer formatting

the first audit must be a 403 with the exact error object. if the target answers differently, the profile update is skipped. both routes must remain on the assigned origin, and the proof comes only from the later audit response. the submitted value is [redacted].

failed approaches and pivots

the first generic capture read clues and route responses but did not perform the guarded update. it therefore never reached the post-update audit evidence. the specific error and paired routes supplied a short, safe decision path, which a fresh regular variant later graded correct.

automation across modes

the audit guard is evaluated anew for each target. the collector uses a separate cookie jar for each mastery target and keeps the proof attached to the question that names that target.

lessons

solve times

public scoreboard · day 23 finish: 1st

  • 3:49.4single#29 of 157 solvers
  • 7:43.2time trial#33 of 152 solvers
  • 8:31.8mastery#26 of 150 solvers
  • #1team day finishof 140 teams that cleared it, in 8:58.5
  • #7single orderof the day's 8, by our flag time

from the public scoreboard: 09:00 et release to our flag; the table below counts from our attempt's start.

the preserved completion record gives assignment start to flag for the single and mastery. it does not preserve the winning time trial's start or wrong-submit history.

mode work to flag attempts wrong submits
single 3m 46.3s from recorded start 1 recorded unknown
time trial unknown unknown unknown
mastery 36.4s from recorded start 1 recorded unknown

the clock

the challenge listed a nominal 120-second timer for ranked timed attempts. the completion record contains no exact deadline, so time left is inferred from the configured timer for mastery.

mode time available time used time left
time trial 120s configured unknown unknown
mastery 120s configured 36.4s from start about 83.6s, inferred

where we got stuck

flags: single, time trial, and mastery [redacted].

{generated by a clanker}