Bounty lifecycle and settlement
A bounty is a workflow with economic consequences. Applications should model the lifecycle explicitly so users can distinguish a proposal, a funded task, a claim, a submission, an approval and an actual settlement.Creation
bounty.create is a funded action for Author, Journalist or Editor roles with an active bond. Creation should establish a clear target, expected work product, funding amount, review path and expiration policy before the user authorizes the action.
If the bounty concerns a Capsule, challenge or evidence gap, store that relationship directly instead of relying on a free-form description that a settlement process cannot verify.
Claim
bounty.claim is a micro-fee action for eligible Author, Journalist, Editor and Validator roles. A successful claim establishes the current assignee.
Read current state immediately before authorization. Two users can view an open bounty at the same time, and one may claim it before the other transaction executes. Treat that as a normal race in shared state.
Submission
bounty.submit is limited to the current assignee and requires an active bond. The submission should reference a verifiable work product whenever possible.
For evidence work, prefer a durable URI plus content hash. For review work, preserve the exact Capsule, claim or evidence item being reviewed. This gives the approval path something concrete to inspect.
Approval
Approval is available to Editor or Validator roles when the actor also satisfies the creator, designated reviewer or protocol authority relationship defined by active policy. Role alone is not universal approval power.Settlement
Settlement is an authorized escrow or treasury operation. Before value moves, the settlement path should re-check:- bounty is in an approvable state
- assignee is still the intended recipient
- required role is still active
- active bond requirement is still satisfied
- treasury safety policy passes
- the bounty was not already paid
- the settlement identifier was not already processed