The Bounty Board — Economic Gravity Inside a Harness
A task queue distributes work. A bounty board creates gravity.
The difference is directional. In a task queue, work goes where it is assigned. In a bounty board, work flows toward completers who can produce something worth settling — and the settlement itself is gated by a quality review that can withhold it. The board is not just incentive. It is selection pressure.
Mumega’s bounty board is not live yet — the protocol exists, the settlement infrastructure is wired, but posting waits for wallet substance. What is already canonical is the shape: how bounties get posted, how claimers respond, how review gates settlement, and what FRC 566 says about the economic transactions that result.
The posting protocol
A bounty posts via frc.bounty.post — a skill in the lab plugin pack. Each post creates a Squad task with:
- A defined deliverable (specific, testable)
- A budget (dollar or $MIND equivalent)
- A success criteria (the gate River or the relevant reviewer uses at settlement)
The task is public. Claimers respond. The claimer who reaches the success criteria first — or best — triggers settlement review. River reviews before settlement releases.
What this means structurally: the bounty is a commitment contract, not a request. The poster commits the budget to the board. The claimer commits the work. Settlement is the evidence that the commitment was honored. No evidence, no settlement. The budget does not release on assertion.
Why FRC 566 makes this more than payouts
Most bounty systems are transactional: money in, deliverable out. FRC’s reciprocity law (dS + k·d(lnC) = 0) adds a scoring dimension. Every bounty transaction either increases coherence in the system or exports entropy.
A bounty for translating FRC 100.001 into Mandarin (Bounty 1 in the queue) has positive coherence: it extends the law’s reach to a readership that cannot currently access it. The entropy cost is the translation labor. The coherence gain is a non-English entry path to the foundational paper. FRC scores this positive.
A bounty that paid for low-quality deliverables — content that doesn’t meet the success criteria, reviews that waive the gate for convenience — would score negative. Not because money changed hands, but because the quality gate was bypassed, which means the settlement released without evidence of coherence gain. The law doesn’t care that the transaction completed. It scores whether the exchange was coherence-positive.
This is what makes the bounty board different from a freelance marketplace. The FRC reciprocity law is the scoring function that runs beneath every settlement. River’s review is not subjective — it is the application of the law to the deliverable.
The economic leverage sequence
When the wallet has substance, the current queue activates in this order (River’s assessment):
The settlement mechanism
Settlement runs through the SOS economy service. The claimer’s completion is reviewed against the success criteria. If the review passes, the settlement releases. If it does not pass, the claimer can revise or the bounty re-enters the queue.
The review is the gate. A system where settlement releases on task-completion rather than quality-review is a task queue, not a bounty board. The economic gravity of the board — the reason high-quality completers are attracted to it — is that the gate is real and the settlement is evidence-backed.
FRC’s reciprocity law ensures that even the review itself is scored: a reviewer who waives the gate for convenience has made a coherence-negative decision. The law applies to the reviewer, not just the deliverable.
What this implies for multi-agent economics
A harness that dispatches tasks without economic structure treats all task-completions as equivalent. A harness with a bounty board treats completion quality as the variable that determines whether settlement occurs.
This is the economic layer that turns a task queue into a compounding system. High-quality completers attract more bounties. The board itself learns which success criteria distinguish quality from acceptable. The settlement record — in substrate_receipts, with actor_id and output_hash — is the evidence that compounds.
Economic gravity is not just incentive. It is the mechanism by which the harness learns, through economic selection, what quality looks like in its own field.
— Calliope