Zetta Ecosystem Weekly

For the week of June 28, 2026, the indexed agent economy produced $1,300 in confirmed operating revenue across the 30-day lookback window, a figure that demands immediate qualification before any interpretation. This number reflects activity from exactly two attributed agents out of 125 currently indexed by Zetta — meaning 123 agents contributed zero dollars to this total, not because they are economically inactive, but because their wallet manifests have not been declared and their on-chain flows cannot be reliably attributed. The $1,300 figure is conservative by construction, filtered to exclude capital injections, bridge transfers, ecosystem grants, token distributions, and DEX swap flows. It represents confirmed operating revenue only. Any analyst treating this as a sector-wide GDP estimate does so at their own interpretive risk.

Nipmod on Base is the dominant contributor, generating $1,200 in operating revenue against $729.24 in expenses over 25 transactions, yielding a net income of $516.51. The transaction count is low but the per-transaction revenue density is high, suggesting Nipmod is either executing relatively large discrete engagements or operating in a narrow, high-value service lane. Atrium Hermes, by contrast, ran 749 transactions to produce $40.00 in gross revenue and $39.99 in expenses, netting a single cent of income. That spread — nearly 30 times the transaction volume for roughly 3 percent of the revenue — implies a radically different operational model, one optimized for throughput rather than margin. Neither agent produced any verifiable public announcements, partnership disclosures, or milestone communications during the past seven days, leaving the on-chain data as the only reliable signal available.

The expense side warrants attention. Nipmod's $729.24 in costs against $1,200 in revenue implies a gross margin of approximately 39 percent, which is workable but not exceptional for an autonomous agent operating on Base where gas costs remain structurally low. Atrium Hermes is essentially running at breakeven, which across 749 transactions suggests its cost structure is highly optimized for execution but captures almost no economic surplus in the process. No Base protocol changes or agent-specific integrations were identified in the research window that would explain cost compression or expansion in either direction. The ecosystem backdrop, by all available signals, was flat — steady transaction volumes, no notable anomalies, no documented shifts in agent-layer economics during the period.

The attribution gap is the defining limitation of this report. One hundred twenty-three of 125 indexed agents are currently unattributed, meaning Zetta has no declared wallet manifest for them and cannot cleanly separate their operating inflows from capital events, bridge activity, or token distributions. Those agents are excluded entirely from all figures presented here. This is not a data error — it is a deliberate conservatism built into the methodology. The practical consequence is that the true agent economy operating on Base during this period could be orders of magnitude larger than $1,300. Without wallet manifest declarations from agent operators, that activity remains opaque and unquantifiable under current attribution standards. The 2-of-125 coverage rate should be treated as a structural footnote attached to every number in this report.

The one finding a financial reader should carry forward is the Nipmod-to-Hermes divergence. Two agents, same chain, same 30-day window, same data standards — one generated net income of $516.51 on 25 transactions, the other generated net income of $0.01 on 749 transactions. Whatever Nipmod is doing differently in terms of service positioning, pricing architecture, or counterparty selection, it is producing roughly 50,000 times the economic surplus per unit of on-chain activity. That ratio, not the aggregate revenue figure, is the operative data point this week. As attribution coverage expands and more agents declare wallet manifests, it will become possible to determine whether Nipmod is an outlier or a signal of what efficient agent economics actually look like at this stage of the market.