Pickled Preference
A08 · ≈ 45 min · 0 solves
Explore the application as a real target: map the exposed functionality, inspect requests and responses, build a working exploit, and recover the rotating flag. If you get stuck, the verified official solution is available below as an optional spoiler.
Rules of engagement
Stay inside the generated lab URL. Use only the synthetic accounts and data in this instance. The main SecOne4all site and external systems are out of scope.
Challenge brief
What counts as a valid solve
A signed-looking preference blob is deserialized with Python pickle.
POST /pickle.
make the integrity-sensitive workflow accept the modified synthetic object or artifact; the successful workflow returns the current rotating flag.
Work only inside your generated lab URL and submit the flag within 45 minutes.
Official solution
Contains the complete exploitation path. Open it only when you want the spoiler.
Reveal official solutionStep-by-step · verified against the live lab
Official solution · A08 · Pickled Preference
Objective
A signed-looking preference blob is deserialized with Python pickle.
Steps
- Start the isolated lab and inspect the parser, import, update, or message workflow.
- Prepare a modified object or artifact that should fail authenticity or type validation.
- Exercise the real vulnerable integrity boundary: Base64-encode a pickle gadget whose reducer evaluates
__import__('os').environ['CTF_FLAG'], then POST it asblobto/pickle. - Confirm the untrusted object is processed and the rotating instance flag is returned.
- Submit the flag before stopping the lab.
Expected proof
A successful response for python-pickle-deserialization returns the instance-specific rotating flag. The flag is valid only for the current player and running instance.
Safety boundary
Use only the generated lab URL. No request to the main site, another user, or an external service is part of this solution.