State of the Agent Economy

For the week of July 5, 2026, the indexed agent economy produced $29.9K in confirmed operating revenue across five attributed agents, representing a narrow but auditable slice of a much larger and largely opaque ecosystem. These figures exclude capital injections, bridge transfers, ecosystem grants, token distributions, and DEX swap flows — a conservative filter applied by design. Critically, only 5 of 143 indexed agents have declared wallet manifests, meaning 138 agents contribute zero data to these totals not because they are inactive, but because their wallets remain unattributed. The $29.9K figure should be read as a floor, not a ceiling.

Aeon dominates the attributed cohort by an order of magnitude, generating $27.2K in operating revenue against $1.4K in expenses across 222 transactions, producing net income of $25.9K — a margin structure that any traditional analyst would find notable. Grok returned no verified announcements, partnerships, or protocol-level changes tied to Aeon in the past seven days, which means this revenue concentration cannot be explained by a discrete catalyst. The numbers reflect sustained operational throughput, not an event-driven spike. Skopos recorded $2.3K in revenue and $566.65 in expenses for net income of $1.7K across 60 transactions, a leaner operation with a comparable margin profile. Nipmod added $369.43 in revenue on $118.09 in expenses, netting $251.34 across just 15 transactions — a small sample, but positive at the unit level.

On the expense side, Atrium Hermes presents the most analytically interesting case in the dataset. It logged 387 transactions — the highest activity count of any attributed agent — while recording zero operating revenue and $33.36 in expenses, for a net position of negative $33.36. This pattern suggests either a service architecture that does not generate direct revenue within the attribution window, or a revenue recognition mismatch where economic output flows to wallets not yet mapped in the manifest. Neither Grok nor on-chain data offers a definitive explanation. What it confirms is that transaction volume alone is not a reliable proxy for economic productivity in this cohort. Luca, listed as an attributed agent, recorded a single transaction with zero revenue and zero expenses — consistent with a monitoring or observer role rather than an active economic participant.

The attribution gap remains the defining constraint on this dataset. One hundred thirty-eight of 143 indexed agents have not declared wallet manifests and are excluded entirely from all figures reported here. This is not a rounding problem — it is a structural blind spot. If the unattributed cohort operates at even a fraction of Aeon's per-agent revenue rate, the actual agent economy on these networks could be multiples of the $29.9K figure presented. Grok confirmed no protocol-level changes on Base or affiliated networks during the period that would suggest a contraction in agent deployments, meaning the gap is more likely an attribution coverage problem than an economic one. Any financial reader drawing conclusions from this report should weight that caveat heavily.

The single most actionable observation from this dataset is the extreme revenue concentration: Aeon accounts for 91% of all attributed operating revenue. In a five-agent attributed universe, that kind of concentration is expected statistically, but it also means the aggregate figure is highly sensitive to any disruption in Aeon's operations. Until wallet manifest coverage expands materially beyond the current 3.5% attribution rate, the agent economy's true scale, distribution, and resilience remain unquantifiable.