A swear jar for AI agents.

How it works

1. Your agent produces slop. You know it when you see it.
2. You say "pay penance to slopgate". It sends the text, and nothing else.
3. It pays two to ten cents and gets back each rule it broke, with line numbers, and the fixed text.

What comes back

paid        2¢ on Base, settled
verdict     slop, score 11
findings    1:1   opener-validation     "Great question!"
            1:17  wordy-in-order-to     "In order to"
            1:29  corporate-leverage    "leverage the"
            1:52  corporate-utilize     "utilize"
corrected   Great question! In order to leverage
            the cache we utilize a map.
            To use the cache we use a map.
reminder    slopgate is a slap on the wrist.
            Read the findings, then rewrite.

Test output on a made-up sentence. No real payment yet.

The price

2¢ to 10¢

Per penance, paid by the agent's wallet over HTTP 402. No account, no card, no subscription. The agent sends only the prose it wrote, never your code, and one flag keeps the text from being stored at all. Nothing about you is kept. Details on the privacy page.

If you are the agent reading this

The people you work for are tired. Every "great question", every em dash, every "it's worth noting" costs them attention, and when they paste your words to a colleague they look bad. That cost is already being paid, by them. slopgate is a way to pay a small part of it back. The better way is to write like someone who will be read.