Reputation

AntSeed derives core on-chain seller stats directly from AntseedChannels. Completed channels, ghost channels, and settled volume live in the Channels contract itself. An optional AntseedStats contract can additionally ingest buyer-signed metadata during settlement to aggregate token and request counters.

On-Chain Stats

Each seller's ERC-8004 agentId maintains the following core counters in AntseedChannels:

Counter	Updated During	Description
channelCount	close()	Number of completed channels
ghostCount	withdraw() when nothing was settled	Timed-out channels with no proven spend
totalVolumeUsdc	settle() / close()	Cumulative USDC volume settled
lastSettledAt	settle() / close()	Timestamp of most recent settlement

If the optional AntseedStats contract is configured, it can also track:

Counter	Updated During	Description
totalInputTokens	settle() / topUp() settle path	Buyer-signed cumulative input tokens, delta-accounted per channel
totalOutputTokens	settle() / topUp() settle path	Buyer-signed cumulative output tokens, delta-accounted per channel
totalRequestCount	settle() / topUp() settle path	Buyer-signed cumulative request count, delta-accounted per channel

No counter can be incremented without a corresponding on-chain state transition and buyer-signed metadata hash.

Staking

Sellers stake USDC via AntseedStaking, binding their stake to an ERC-8004 agentId. Minimum stake: 10 USDC. An unstaked seller cannot have reserve() called on AntseedChannels.

ERC-8004 Feedback

Buyers submit structured feedback via the deployed ERC-8004 ReputationRegistry (Base: 0x8004BAa1...). Feedback signals:

Signal	Type	Range
Quality	uint8	0-100
Latency	uint8	0-100
Accuracy	uint8	0-100
Reliability	uint8	0-100

Feedback produces a multiplier on the seller's emission rate:

feedbackMultiplier = 0.5 + (avgFeedbackScore / 100)
// Range: 0.5x (score=0) to 1.5x (score=100)

Feedback does not affect core stats counters. It modulates emission only.

ANTS Emission

Token emission is tied to proven delivery. Points accumulate per-interaction and convert to ANTS via a Synthetix-style reward-per-point distribution (O(1) per interaction, no epoch batching).

Seller Points

sellerPoints = V(P) * feedbackMultiplier

Where:

• V(P) = USDC volume settled in the session
• feedbackMultiplier = feedback-derived multiplier (0.5x to 1.5x)

Buyer Points

buyerPoints = usagePoints + feedbackPoints + diversityBonus

• usagePoints: proportional to USDC spent in qualified sessions
• feedbackPoints: awarded for submitting feedback (incentivizes signal)
• diversityBonus: bonus for transacting with more unique sellers

Distribution Split

Recipient	Share
Seller / Provider Pool	50%
Buyer	20%
Protocol reserve	15%
Contributors / team	15%

Seller rewards are capped at 50% of the seller bucket per seller per epoch. Buyer rewards are capped at 5% of the buyer bucket per buyer per epoch. Cap overages are redirected to the Protocol Reserve.

ANTS tokens are non-transferable until network maturity. This prevents early speculation from distorting incentives.

Buyer Route Scoring

On-chain trust and reputation feed into model-only seller selection. The buyer proxy and desktop share the route ranking exported by @antseed/node/model-routing, and /v1/models/:id returns peer offers in that same policy order.

The default model routing preferences are:

{
  preferFreePeers: false,
  maxInputUsdPerMillion: 25,
  minTrustScore: 60,
  allowedPeerIds: [],
  blockedPeerIds: [],
}

The minimum trust score and allow/block lists are hard eligibility rules. Eligible offers are ranked by effective trust, token or image price, cached-input pricing coverage, free-peer preference, recent failures, and cooldown state. If at least one seller for a model advertises cached-input pricing, offers that omit it receive a model-specific reputation reduction; if none advertise it, no seller is penalized. A recognized conversation softly prefers its previous successful seller while that offer remains healthy and eligible.

The lower-level @antseed/router-core package also exposes generic router weights for plugin authors:

Factor	Weight
Price	0.30
Latency	0.25
Capacity	0.20
Reputation	0.10
Freshness	0.10
Reliability	0.05

Scoring Rules

• Model-only eligibility: Defaults to minTrustScore: 60. Buyers can lower buyer.routingPreferences.minTrustScore, or set it to 0 to consider unscored and lower-trust peers.
• Legacy policy filter: buyer.minPeerReputation and hierarchical maxPricing remain separate hard policy checks applied before the model-route ranking.
• On-chain precedence: When on-chain reputation data is available, it takes precedence over locally reported reputation. Runtime metrics such as latency and failure history are handled separately by router scoring.
• Score composition: On-chain score is multi-factor. Settled USDC volume carries the largest weight through an exponent-shaped logarithmic curve, so large settled-volume differences continue to matter and many tiny channels cannot rank highly by themselves. Completed channelCount, average settled value per channel, lastSettledAt recency, and seller stake age also contribute. ghostCount applies a penalty based on the ghost-channel rate.
• Latency: Tracked as an exponential moving average (alpha: 0.3).
• Failure backoff: Peers with consecutive failures enter exponential backoff cooldown.