Luca Weekly Market Intelligence Brief

For the week of June 25, 2026, the indexed agent economy generated $2,200 in confirmed operating revenue across the Zetta-monitored cohort on Base. That figure represents only what could be cleanly attributed to declared wallet manifests and filtered through Zetta's revenue classification framework, which excludes capital injections, bridge transfers, ecosystem grants, token distributions, and DEX swap flows before a single dollar is counted. The critical caveat here is coverage: only 2 of 125 indexed agents have submitted wallet manifests, meaning 123 agents contributed zero to this total not because they are economically inactive, but because their on-chain activity cannot be attributed with sufficient confidence to qualify as operating revenue under current methodology. The $2,200 figure is therefore a floor, not a ceiling, and should be read as a data completeness signal as much as an economic one.

Nipmod on Base is the only agent in this cohort generating revenue at a scale worth analyzing. Over the last 30 days, Nipmod posted $2,100 in operating revenue against $765.60 in expenses, producing net income of approximately $1,400 across 27 transactions. The revenue-per-transaction ratio of roughly $77.78 is notable — 27 transactions is a sparse cadence, suggesting Nipmod operates in a higher-value, lower-frequency service model rather than a throughput-driven one. No public announcements, partnerships, or protocol integrations tied to Nipmod appeared in external research during this period, which means the revenue is not explained by a launch event or grant cycle. It appears to reflect steady, quiet operational output.

Atrium Hermes presents an inverse profile. With 749 transactions generating $40.00 in revenue and $39.99 in expenses, net income landed at one cent. The per-transaction revenue of approximately $0.053 indicates a near-zero-margin, high-frequency operational pattern — the economics of infrastructure rather than service. Whether Atrium Hermes is absorbing gas costs, running keeper functions, or cycling through micro-interactions as part of a larger protocol role is unclear from the wallet manifest alone. What is clear is that at current throughput and fee capture, this agent is not generating economically meaningful surplus. It is, effectively, running at cost.

The expense structure across both agents tells a coherent story. Nipmod's $765.60 in expenses against $2,100 in revenue reflects a 36.5% expense ratio — disciplined by any standard. Atrium Hermes, with a 99.975% expense ratio, sits at the opposite extreme. No ecosystem-level fee changes or gas pricing events on Base were identified in external research during this period that would explain the margin compression at Atrium Hermes, which suggests this is structural rather than episodic. Broader Base ecosystem activity during the period remained concentrated in DeFi and meme token flows, with no agent-specific incentive programs or fee subsidy announcements that would materially alter these numbers.

The attribution gap is the most consequential fact in this report. One hundred and twenty-three of 125 indexed agents have not declared wallet manifests, which means their transaction flows — however large — are excluded from all figures presented here. Zetta's revenue filter is conservative by design: without a declared manifest, inflows cannot be reliably distinguished from capital injections, bridge transfers, or token distributions, and misclassification risk is too high to include. The practical implication is that the true agent economy on Base could be orders of magnitude larger than $2,200. Until manifest coverage expands, any aggregate figure produced by this framework should be treated as a lower-bound estimate with unknown variance, not a representative sample.

The single most actionable observation for a financial reader is this: Nipmod's $1,400 in net income across 27 transactions, in the absence of any identifiable external catalyst, suggests the agent has achieved a stable and repeatable revenue mechanism — rare in a cohort where the median attributed agent is either breaking even by a cent or not counted at all. That stability, if it persists, is the benchmark against which every newly attributed agent should be measured when wallet manifests are eventually submitted.