
Yebo is an evidence layer designed for machine-originated payments, focusing on ensuring that AI and software agents cannot execute sensitive financial actions without verified human approval. It addresses the challenge that traditional disbursement controls, which were built for human workflows, are not sufficient to provide durable, auditable evidence for payments initiated by software. Yebo ensures that money cannot leave a company without cryptographic proof of an authorized human's approval, creating artifacts that auditors can verify years later.
The platform operates through a three-stage pipeline for every sensitive action: intent declaration, policy verification, and authorized execution. In the first stage, an agent submits a structured, signed proposal detailing the intended action, parameters, and identity. Next, Yebo evaluates this intent against a defined policy bundle, enforcing deterministic rules, scoped permissions, threshold checks, and dependency graphs in under 50 milliseconds. Only after successful verification does the action proceed to execution, carrying a signed receipt that records what was approved, under which policy, and by which agent. This process is designed to be deterministic, fail-closed (actions are denied if any ambiguity or policy unavailability arises), and produces immutable audit trails for every decision.
Yebo enforces runtime policy checks inline at the call site, allowing instant propagation of updates across agents and regions. Each verification generates a signed, replayable, and tamper-evident receipt, ensuring traceability and compliance. Permissions are tightly scoped and time-bound, never granting blanket access to agents. The system is built to deny execution by default if policies are unreachable or ambiguous, ensuring no implicit approval paths exist.
The tool is positioned for scenarios where software can move money, such as vendor payments, fund transfers, and payouts, particularly those above a certain threshold. It is intended for organizations seeking to implement mandate controls in automated payment workflows, providing proof of human approval before execution. Yebo's features are aimed at developers and finance teams responsible for securing machine-originated disbursements and maintaining compliance with audit requirements.
In the Infrastructure & Backend space, Yebo takes a focused approach. It focuses on securing and authorizing sensitive actions performed by AI agents at runtime with deterministic policy enforcement. It is built as a B2B product for developers building AI agent systems. It runs on API.
Yebo first shipped in 2026. Key capabilities include policy evaluation, runtime authorization, and audit logging. It exposes integrations via a public API.
Latest indexed changes and source events
Trust does not scale at machine speed verified by the PulseGate indexer
Other apps tracked under the same category.