aequai ~/agent-operations-safety-audit · ai evidence operations book ↗
aequai ~ / agent-operations-safety-audit
$ aequai page --agent-operations-safety-audit

Review agent authority before it grows.

This audit maps what agents can read, draft, write, send, deploy, mutate, approve, or escalate, then defines checkpoints before risky authority expands.

// what we review: Agent and tool inventory, action classes, permissions, logs, rollback paths, exception handling, rejected actions, and human checkpoints by risk tier and output destination.

// What you receive: Typed artifact map, operation boundary, provenance/evidence map, breaker tests, gate record, and transition recommendation. No exploit testing, credential handling, production operation, or autonomous customer messaging is implied.

$ aequai contact --book diagnostic

Map one workflow first.

Bring one workflow, one team, or one evidence problem. The useful next step may be proceed, shrink, park, or stop.