A supply chain is already a network of parties that share no systems and trust each other conditionally. Agent-to-agent coordination makes that faster, and makes unverified delegation considerably more expensive.
Source: 2025 industry surveys of enterprise AI agent adoption.
Shipper to broker to carrier to port to drayage to warehouse. Six organizations, six systems, no shared identity layer — and now each is putting agents in front of tendering, appointment booking, exception handling and status reconciliation.
Each handoff is cross-org by definition, so there is no internal governance that reaches it. EDI and API keys authenticate the connection. Neither says anything about whether the party on the other end performs.
Carrier scorecards exist, but they are private, per-shipper, backward-looking and unavailable at the moment of tender — which is the only moment they would change a decision.
Assay gives every participant's agent a portable identity and a trust score built from outcomes their counterparties confirmed. It is readable at tender time by one call, and it is not locked inside any one shipper's TMS.
The logistics preset tunes the model for the sector, where volume is high and recency matters more than deep history. Capabilities score independently, so a carrier proven on drayage is not automatically trusted with a temperature-controlled lane.
Attestations flow both ways. A carrier's agent builds a record that follows it to the next shipper, which turns good performance into a portable asset rather than a private note in someone else's system.
A score you can read during tender rather than a scorecard reviewed next quarter.
A carrier's record follows its agent. Good performance becomes an asset the carrier carries, not a private note in one TMS.
Drayage, reefer, hazmat and appointment booking score independently. Proven on one lane type is not proven on all.
Parties that share no EDI, no TMS and no prior relationship can still verify each other. The identity is public; the operational data is not.
Discovery is how you find a qualified alternative when the primary declines.
verified reflects the trust
threshold and account status. It does not fall to false because a claimed capability
is unproven — an unproven capability simply does not appear in
capabilities_confirmed. Gate on both fields.It complements it. Your scorecard captures commercial history specific to you. Assay captures cross-shipper, counterparty-verified outcome history that is available at the moment of tender.
The incentive is on their side: a portable record that follows them to the next shipper. Registration is free and takes one call, so the cost of participating is close to zero.
Yes. Capabilities are arbitrary strings scored independently, so drayage, reefer and hazmat can be separate claims with separate thresholds.
Cold start returns a neutral score and no confirmed capabilities, which fails a real threshold. Start them on lower-stakes freight to build record, or let them import verified prior work.