New release v2026.626.0: Hermes agents built in, task watchdogs, and ask work mode.
See what's newReporting agents assemble the numbers on schedule, every agent operates under a hard budget cap, and every task carries a receipt. Finance gets the visibility other tools promise.
Each agent operates under a budget with a hard cap — at the limit, work stops and a top-up becomes an approval decision. Spend by agent, team, model, and goal is live, so the monthly AI number is a glance, not a project.
Publish “Comparing agent runtimes” to the blog
Reporting agents pull the numbers, build the pack, and deliver it on close cadence — with sources cited and anomalies flagged. The analyst reviews and signs instead of copying cells at midnight.
Market brief: what changed since Monday
— Competitor X shipped usage-based pricing [source]
— Two enterprise threads mention agent budgets [source]
— Regulatory draft affects EU deployments [source]
3 claims verified · 1 refuted and omitted
Each figure in a report links back to the run that produced it: inputs read, steps taken, cost incurred. Audit season becomes navigation, not excavation.
“Nobody should discover AI spend in arrears. Caps are enforcement, receipts attach to the work itself, and the monthly number is a glance — that is the bar we build to.”
Caps, receipts, and reporting packs — from day one.