Spend logs

Every payment — approved or blocked — produces a structured, human-readable log entry. Each entry records the agent, service, amount, reason, result, and the exact policy rule that decided the outcome.

log-entry.json

{
  "agent": "researchbot-01",
  "service": "risk-report-api",
  "amount": "0.03",
  "asset": "USDC",
  "reason": "Fetch risk report",
  "result": "approved",
  "rule": "under_max_payment",
  "latencyMs": 94,
  "ts": "2026-06-10T09:31:04Z"
}

Each log entry is also viewable as a receipt with its full lifecycle status.