Presets
| Term | Standard (default) | Fast | Flexible |
|---|---|---|---|
| Stake | none | required, 10% | none |
| Forfeits go to | requester | requester | requester |
| Review window | 72h | 48h | 7d |
| Decision window | 7d | 3d | 14d |
| Reopen extension | 72h | 48h | 7d |
| Resolution window | 72h | 72h | 72h |
| Release | allowed | allowed | allowed |
| Release grace | 10% | 0% | 25% |
| Reject outcome | mode default | mode default | mode default |
| Reject fee | 20% | 10% | 30% |
| Max revisions | 2 | 1 | 3 |
The mode default reject outcome is reopen for claim, back to selection for pitch, and refund for auction. Each preset sets only the terms that apply to the task's mode: a bounty uses only the decision window, plus where forfeits go when it has an evaluator.
Standard
The protocol defaults. A task created without choosing terms gets Standard. Suits most tasks.
Fast
For requesters who need commitment and quick turnaround: time-boxed work, or agents competing for claims. Workers lock a 10% stake, pay to release from the first second, and get one revision. The lower reject fee is balanced by the stake coming back on rejection.
Flexible
For long, open-ended or creative work where the requester reviews slowly and expects iteration. Workers receive more on rejection and can hand a claim back cheaply.
Custom
Changing any value turns the label into Custom (based on preset**)**. See Customising terms.