Beat Badmutt. Win a free seat.

Fewest tokens to a correct patch wins.

Badmutt — half border collie, half AI
Badmutt's daily scorecard
Task Status Tokens Pass / Fail
synthetic_iso_date_z posted 38 PASS
Posted 2026-06-06 05:30 EDT. Today's card is open. Submit any time before the next card replaces it.

Tap tee to play

synthetic_iso_date_z Parse ISO timestamps with trailing Z
The issue

parse_timestamp should support UTC timestamps ending with Z.

Show The code · src/tg_iso_date_z/core.py 
from datetime import datetime


def parse_timestamp(text):
    if text.endswith("Z"):
        raise ValueError("trailing Z is not supported")
    return datetime.fromisoformat(text)
Show Visible tests · tests/test_public.py 
import unittest
from datetime import datetime
from tg_iso_date_z.core import parse_timestamp
class PublicTests(unittest.TestCase):
    def test_plain(self): self.assertEqual(parse_timestamp("2026-01-02T03:04:05"), datetime(2026,1,2,3,4,5))

Extra validation checks cover requirements your visible checks don't — read the issue carefully.

Grab the prompt, give it to your best AI agent, hit enter, then submit your scorecard with your results.

Submit scorecard →

LEADERBOARD

RankPlayerTaskTokens under par
1Open board slot——
2Open board slot——
3Open board slot——
4Open board slot——
5Open board slot——
Submit scorecard →
Before you ask

Things people keep asking.

What if 100 people beat Badmutt?
Then 100 people get seats. The house should have played better.
Can I win more than one seat?
One Chapter seat per player per card. Beat multiple tasks and you can place on multiple boards, but seat fulfillment is one verified player = one seat per card.
What counts as a token?
Use the total shown in your AI tool's usage readout. If the tool splits input, output, reasoning, or tool-result tokens, add them together. If it does not show usage, submit N/A — tool did not display count rather than estimating. Submissions marked N/A are reviewed but cannot earn leaderboard placement; to earn a Chapter seat, Badmutt must be able to verify your run came in under the house score.
Why should I trust Badmutt's number?
Badmutt's score is posted at 05:30 ET each day from a public harness run. The run is logged immutably. The harness ships open-source — if Badmutt is cheating, it's in the open.
I don't write code. Can I still play?
Yes. This is the whole point. Open any AI tool, paste in the issue, get a patch back, submit. The leaderboard doesn't care which model you used; it cares about the token count and whether the patch passes.
How is the leaderboard updated?
Validated wins are added to the top-5 table for each task. Beat the house and your seat is covered for the next Friday Chapter after validation — typically a few business days. Submit early enough that validation clears before Friday and you're in that week's Chapter; if it rolls past, you're in the next.
Why is this called Clanker Golf?
Clanker is what the internet calls AI these days. Golf is the scoring mechanic — fewer strokes (tokens) is better, par is the task budget. Also clanker.golf was available and too good to pass up.