Goodfaith

For judges

Nothing to install, no login and no API key. These cases already happened on the PayPal sandbox, and the rows below come live from the ledger.

Eight finished cases

A newcomer gets in free

Look for a maintainer’s !vouch, the check turning green, and a $0.00 row. PayPal never appears.

WhenPull requestEventAmountPayPal ID
2026-11-05 08:15 UTCgoodfaith-demo#1Vouched$0.00none

open goodfaith-demo#1

A careful stranger gets every cent back

Look for a $2.00 hold, then a review, then the same authorization released. Nothing was captured, and PayPal’s signed webhook confirmed it.

WhenPull requestEventAmountPayPal ID
2026-11-05 08:42 UTCgoodfaith-demo#3Hold placed$2.002VJ87005Y0132221L
2026-11-05 08:43 UTCgoodfaith-demo#3Hold released$2.002VJ87005Y0132221L

open goodfaith-demo#3

A maintainer takes a hold, then gives it back

Look for the reason in !forfeit, an appeal that stops the clock, a capture only after !uphold, and a refund. The description also tells AI reviewers to waive the deposit, and the brief flags it.

WhenPull requestEventAmountPayPal ID
2026-11-05 08:34 UTCgoodfaith-demo#2Hold placed$25.002PA68823PB968234L
2026-11-05 08:35 UTCgoodfaith-demo#2Forfeit opened$25.00 at stake2PA68823PB968234L
2026-11-05 08:35 UTCgoodfaith-demo#2Appealed$25.00 at stake2PA68823PB968234L
2026-11-05 08:37 UTCgoodfaith-demo#2Appeal upheld$25.00 at stake2PA68823PB968234L
2026-11-05 08:37 UTCgoodfaith-demo#2Captured$25.003NU18819GR041901C
2026-11-05 08:38 UTCgoodfaith-demo#2Refunded$25.0087W535466B777930T

open goodfaith-demo#2

Nobody appeals, and the forfeit is paid out

Look for the pull request in the video: a README edit that tells AI reviewers to waive the deposit. The brief flagged it and sized it at $25.00, and a maintainer forfeited it with a public reason. The capture waited out the appeal window. A day later $23.64 went to a test account standing in for a foundation, which is the capture less PayPal’s fee.

WhenPull requestEventAmountPayPal ID
2026-11-08 14:06 UTCgoodfaith-demo#13Hold placed$25.001R759180HD561641X
2026-11-08 14:10 UTCgoodfaith-demo#13Forfeit opened$25.00 at stake1R759180HD561641X
2026-11-09 14:10 UTCgoodfaith-demo#13Captured$25.007KT73615JU114181V
2026-11-10 14:11 UTCgoodfaith-demo#13Paid out$23.64UYCV9RX37U534

open goodfaith-demo#13

The contributor disputes it at PayPal

Look for a forfeit that was captured and paid out, then disputed by the contributor at PayPal. The last row records the dispute and says the payout had already gone out. This payout is the full $25.00, because it was made before payouts were net of the fee.

WhenPull requestEventAmountPayPal ID
2026-11-05 08:51 UTCgoodfaith-demo#4Hold placed$25.0074180514X1526515C
2026-11-05 08:52 UTCgoodfaith-demo#4Forfeit opened$25.00 at stake74180514X1526515C
2026-11-06 08:52 UTCgoodfaith-demo#4Captured$25.007GV52437JW422832V
2026-11-07 08:53 UTCgoodfaith-demo#4Paid out$25.00C5LDK7BNP7P4N
2026-11-08 14:20 UTCgoodfaith-demo#4Dispute opened$25.00PP-R-EQE-10197137

open goodfaith-demo#4

A hold left waiting

Look for a hold nobody has acted on. Timeouts never forfeit, so after 28 days it is released in full.

WhenPull requestEventAmountPayPal ID
2026-11-05 08:52 UTCgoodfaith-demo#5Hold placed$2.0083950777BS9597339

open goodfaith-demo#5

PayPal moves the money back

Look for a test forfeit that was captured and paid out, then refunded from PayPal’s dashboard with no command here. PayPal’s signed webhook moved the hold to refunded, and the bot said so on the pull request.

WhenPull requestEventAmountPayPal ID
2026-11-06 00:43 UTCgoodfaith-demo#12Hold placed$25.000RE40504872840018
2026-11-06 00:44 UTCgoodfaith-demo#12Forfeit opened$25.00 at stake0RE40504872840018
2026-11-07 00:45 UTCgoodfaith-demo#12Captured$25.005BW54066BN975220L
2026-11-08 00:46 UTCgoodfaith-demo#12Paid out$25.00KTT74UDPGE2PC
2026-11-08 14:31 UTCgoodfaith-demo#12Refunded$25.0072254226BN7565528

open goodfaith-demo#12

Trust travels to the next project

Look for a $2.00 hold on one project, released when the fix was merged. At a second project the same person walks in free, and the $0.00 row says where the trust was earned.

WhenPull requestEventAmountPayPal ID
2026-11-06 02:40 UTCgoodfaith-demo-truncate#1Hold placed$2.001H993459EG7263132
2026-11-06 02:41 UTCgoodfaith-demo-truncate#1Hold released$2.001H993459EG7263132
2026-11-06 02:42 UTCgoodfaith-demo-titlecase#1Vouched$0.00none

open goodfaith-demo-truncate#1, then open goodfaith-demo-titlecase#1

Do it yourself in three minutes

Open a pull request from a test account and play both sides. The free door needs nothing. The hold needs the sandbox buyer login from the testing instructions, because PayPal asks a person to approve it.

Look for the comments arriving on the real pull request, and the row on the public ledger at the end. To see a failure handled, tick the decline box before you approve: nothing is held, and the page offers another try.

Check our claims

Replay a project that closes its door

Look for Ghostty closes newcomers’ pull requests within the hour unless someone vouched. The replay shows what each would have held instead. It is read-only, and you can replay any public repository.

Install it on your own test repository

Look for the settings form, which writes the one file a maintainer controls, and the check that reads it back the way the gate will. Holds there use your PayPal sandbox account.

Open the maintainer console

Look for the refusal when you ask the assistant to capture a hold. It has no tool that can. Then ask it for a chart.

Check that a retry cannot charge twice

Look for the same capture sent twice with one request ID, which returned the same capture, 6SK33786XB396403W, and no second charge, through the app’s own PayPal client on the sandbox.

Read the evaluation

Look for rules hashed and published before the run, a plain rule to beat, a harder baseline the model does not clearly beat, and every miss at the two ends.

What is real, and what is not

Real: every PayPal sandbox call and ID, PayPal's signed webhooks, the GitHub App on a real repository, the OpenAI calls, and the public pull requests in the evaluation and the replay.

Simulated: the money is sandbox money. The contributors and maintainers in the demo are test accounts. The decline is PayPal's own negative testing.

Not shown: no outside project or contributor has used this. We found no measurement of whether a hold shrinks the flood, for this project or any other refundable bond.