Auto-Approval & AI narration
When you're ready, a single Guardrails toggle lets low-risk, in-policy, high-confidence actions skip the manual gate. The engine still decides deterministically; AI only writes the plain-English summary.
Auto-approval is a confidence-and-Guardrail gate plus a toggle you control — not a model judging whether to spend. The toggle simply lets actions that already pass the engine's deterministic confidence and Guardrail filters skip the human approval step. High-risk categories still require your approval.
Every account starts in Shadow Mode with zero execution. You climb the Autonomy Ladder — Manual Gate, Guided Execution, Trusted Automation, Intraday Autonomy — only by opting in, and the higher rungs require a track record of clean Reconciliation runs first.
AI explains Manifests, writes your daily digest, and helps with onboarding copy. It never decides bids, budgets, pauses, or negatives, and it never sees raw account data it doesn't need.
Auto-approval is opt-in, scoped, and reversible, and the emergency stop is always one tap away. Circuit breakers — spend spikes, drift, API errors — can auto-pause and alert you on the higher stages.
At a glance
Get started
Start in Shadow Mode for free, or take Hosted Starter for a 30-day trial — approve from Telegram, execute safely, reconcile everything.